A Structural Model of Personality Factors, Learning Approaches,
Thinking Styles and Academic Achievement
KADIVAR, PARVIN & SHOKRI, OMID
Department of Psychology
Tarbiat Moallem University
NO 49, Mofateh Ave, Tehran
IRAN
Abstract: - In This Research a short form of the Big Five personality factors Inventory, Study Process Questionnaire, and Thinking Styles Inventory were conducted on a sample of 419 subjects (214 and 205 males and females respectively). Measures of academic achievement were also obtained. Using the path analysis, the direct and indirect effects of personality traits on learning approaches (deep and surface approaches), thinking styles (thinking styles have five dimensions – Functions, Forms, Levels, Scopes, and Leanings) and academic achievement were tested. The obtained results indicated that openness and conscientiousness have a significant positive effect on deep learning approach, while openness and conscientiousness have significant negative effect and neuroticism has a significant positive effect on surface learning approach. The results also showed that openness, neuroticism, conscientiousness, and extraversion respectively have significant positive effect on judicial, global, and liberal thinking styles; on executive, local, and conservative thinking styles; on legislative and hierarchical thinking styles; and on global, liberal, hierarchical and external thinking styles (with the exception of openness that has a significant negative effect on conservative thinking styles).
Key-Words: -Structural Model, Personality Factors, Learning Approach, Thinking Style, Academic Achievement.
1 Introduction
The role of thinking styles in the academic achievements have been examined in different cultures such as Hong Kong china, Philippine, Spain, and the United States. Zhang (2002) indicated that conservative thinking style and global and liberal thinking styles are negative predictor for students' academic achievement. The results of Zhang (2001, 2002) indicated that conservative and hierarchical thinking styles have a positive correlation and creativity generating styles such as judicial and legislative thinking styles have negative correlation with academic achievement. Bernardo et al. (2002) concluded that conformity based thinking styles and respect to authorities is positively related with academic achievement. There was also a positive correlation between judicial thinking style and academic achievement. Cano-Garcia and Hallgesh (2000) indicated that students with internal and executive thinking styles have a better academic achievement.
The idea of relation between the people preferred method in learning with personality factors in not new. For example Messick (1984) as cited by Duff et al. (2003) pointed out that the personality factors determine the formation of learning approach or people preferred method in processing information. Some of other researchers pointed that learning styles or approaches is a sub set of personality (such
as Sternberg) or learnt components of personality (Furnham A., Jackson, C.J & Miller T 1999).
Zhang (2003) conducted a research to answer this question that whether the Big Five personality factors could predict learning approaches. The results of this research indicated that factors such as openness and conscientiousness are strong predictors and can explain student's differences in using learning approaches than other factors. On these bases, conscientiousness and neuroticism could be considered as good predictors for deep and surface approaches, respectively. Duff et al. (2003) reported that: structural equation modeling identifies the Big Five personality factor scores account for between 22.7% and 43.6% of the variance across scores on the three approaches to learning dimensions. A linear regression analysis with academic achievement as the dependent variable and age, prior educational attainment and conscientiousness as independent variables, accounted for 24.1% of the variance in academic achievement.
In order to predict the probability of available relationships between the research variables and academic achievement in this study first a model was selected on the basis of literature review, research background, and research findings; second, after assessing the relationships between the research variables in framework of a casual model, the significance of each variable were assessed and the coefficients were estimated, and finally our proposed model was fitted to the data. The baseline model that we were trying to fit it to our data is provided at the following figure.
hypotheses: Personality factors including openness, conscientiousness, and extraversion; deep learning approach; and thinking styles including judicial, legislative, global, hierarchical, and liberal have a direct
as well as positive effect on academic achievement. And personality factors including neuroticism; surface learning approach; and thinking styles including executive, local, and conservative have a direct and negative effect on academic achievement.
Personality factors including openness, conscientiousness, and extraversion through deep learning approach and thinking styles including judicial, legislative, global, hierarchical, and liberal have indirect
as well as positive effect on academic achievement. And personality factors including neuroticism through surface learning approach and thinking styles including executive, local, and conservative have a direct as well as negative effect on academic achievement.
2 Methodology
2.1 Participants
Subjects consisted of 419 students (214 and 205 male and female respectively) at the Tarbiat- Moalem University in Tehran that
were selected based on multi-stage cluster sampling.
2.2Instruments The Big Five Inventory (BFI): To measure the Big
Five personality dimensions, we used the BFI or the big five personality factors Inventory: A short form of the Big Five personality factors questionnaire is a paper- pen scale that has 44 items and has been designed by Sanjay Srivastava and Oliver P. John in 1991. BFI items are rated on a 5-point scale ranging from 1= disagree strongly to 5= agree strongly. The 44 BFI items consist of short and easy-to-understand phrases to assess the prototypical traits defining each of the Big Five dimensions. In the present research the coefficient alpha reliabilities of the five scales were computed and reported as follows: 0.83 for Conscientiousness; 0.55 for Agreeableness; 0.79 for Neuroticism; 0.79 for Openness; and 0.64 for Extraversion.
A. The Biggs’ Study Process Questionnaire (SPQ) was developed by Biggs (1987) as a 42-item self-report inventory designed to evaluate students’ approaches to learning (SAL) in the higher education context. The SPQ conceptualizes SAL in terms an equal composite of a motivation and strategy sub-scale in three broad learning approach scales (deep, surface and achieving approach). The original SPQ consists of 42 items grouped into 6 sub-scales, these being: surface motivation (SM), surface strategy (SS), deep motivation (DM), deep strategy (DS), achieving motivation (AM) and achieving strategy (AS). Students are asked to respond to each item by indicating how true the statement is of them using a 5-point Likert scale, from 1= “this item is never or only rarely true of me” to 5 = “this item is always or almost always true for me”. Biggs et al. (2001) recognised the need for a two-factor version of the SPQ and developed a revised and shortened version of the instrument: a two-factor model of a deep approach and a surface approach. This version was evaluated with 495 undergraduate students through the use of confirmatory factor analysis using AMOS. The internal consistency of the sub-scales as measured by the use of Cronbach alpha are quite acceptable (values of α = 0.79 for deep approach and α = 0.77 for surface approach). The two models having the best fit to the data are represented by the original six sub-scale model which loads onto two higher order factors, a deep approach and a surface approach to learning, or a seven sub-scale version which loads on the same two higher order factors.
B. Thinking Styles Inventory: Thinking Styles Inventory (TSI) is a paper pencil questionnaire that was designed by R. J. Sternberg and R. K. Wagner's (1992) based on the theory of mental self-government (MGS). The Thinking Styles Inventory is a self-report inventory. The whole inventory is designed for measuring 13 types of thinking styles, consisting of 65 items (statements). Each item belongs to one thinking style and 5 items assess one style. The participants rate themselves on a 7-point Likert scale. In the present study, the Cronbach’s alpha coefficients were used to estimate the reliabilities of the 15 styles. The alpha coefficients were respectively for the legislative style= 0.74, executive style= 0.68, judicial style= 0.71, global style= 0.71, local style= 0.58, liberal style= 0.85, conservative style= 0.82, hierarchical style= 0.75, monarchic style= 0.80, oligarchic style= 0.81,
anarchic style= 0.53, internal style= 0.71, and external thinking style= 0.82.
An inventory was used to gather data on students' academic achievement such as grade point average. When the participants were filling the above inventories they were asked to write down their ID number and academic division code which are usually displayed on the student ID card in the above section of each inventory. Data on academic achievement were gathered based on the ID number and academic division code from the registrar's office of their respective faculties.
3 Procedure
After conducting the designed scales, and gathering the required data, the raw data were entered into a computer file. Then the baseline model which had already been depicted and other models were fitted to the data.
The research design we used in our study leads us to call it as a co relational study which is an example of descriptive research. Here to investigate the role of personality factors, learning approaches, and thinking styles in the academic achievement we used the path analysis method. There were a lot of questions to answer in the questionnaires, so data had to be collected carefully. If values were missing, we couldn't get adequate information of the variables.
4 Results
The results of statistical analysis are provided at the below table. The information of this table indicates that from among the personality factors (exogenous variable) factors such as openness, neuroticism, conscientiousness, and extraversion; and from among thinking styles (independent endogenous variable) dimensions such as legislative, liberal, conservative, hierarchical, and external thinking styles; and deep and surface learning approaches (independent endogenous variable) are positively and significantly correlated with academic achievement (dependent endogenous variable). Further more, from among thinking styles variables (independent endogenous variable) legislative, judicial, global, liberal, conservative, hierarchical, internal, and external thinking styles are positively and significantly correlated with openness; legislative, executive, global, local, liberal, hierarchical, monarchic, internal, and external thinking styles are positively and significantly correlated with neuroticism; legislative, executive, judicial, global, local, liberal, hierarchical, monarchic, internal, and external thinking styles are positively and significantly correlated with conscientiousness; legislative, global, internal, and external thinking styles are positively and significantly correlated with agreeableness; legislative, judicial, global, liberal, conservative, hierarchical, monarchic, and external thinking styles are positively and significantly correlated with extroversion; legislative, executive, judicial, local, conservative, and hierarchical thinking styles are positively and significantly correlated with surface learning approach (independent endogenous); And legislative, executive, judicial, global, liberal, conservative, and hierarchical thinking styles are positively and significantly correlated with deep learning approach (independent endogenous).
Using the path analysis, the hypotheses of this research namely the predicative role and the direct and indirect effects of (1) personality traits; (2) learning approaches (deep and surface), and (3) thinking styles (function, form, level, scope, and learning dimensions) on academic achievement were tested.
Model I
The effects of exogenous variables (personality factors) and independent endogenous variables (deep and surface learning approaches, and function dimension of thinking styles) on dependent exogenous variable (academic achievement) indicated that the model account for 55% of the variance in academic achievement. From among the
exogenous variables just factors such as agreeableness, neuroticism, conscientiousness, and extroversion had a significant effect on academic achievement. Their regression coefficients (beta) were +0.16, -0.14, +0.12, and +0.11 respectively.
The regression coefficients of research's exogenous variables including openness and conscientiousness with deep learning approach respectively were equal
to +0.36, and +0.20; and the regression coefficients
of openness, neuroticism, and conscientiousness with surface learning approach respectively were equal to -0.22, -0.21, and -0.29. Thus it can be concluded that they have a significant effect on learning approaches (figure 2).
Also, the regression coefficients of openness, neuroticism, and conscientiousness with judicial, executive, and legislative thinking styles were equal
to +0.15, +0.11, and +0.12 respectively; therefore they have a significant effect on these styles.
The regression coefficients of research independent endogenous variables including deep and surface approaches with judicial, executive, and legislative thinking styles (the dimension of functions) were equal to +0.15, -0.13, and +0.11 respectively; therefore they have a significant effect
on the dimension of functions. The regression coefficients of judicial, executive, and legislative thinking styles (the dimension of functions) with academic achievement were equal to +0.30, -0.21, and +0.24; therefore the dimension of functions has
a significant effect on academic achievement.
The exogenous variables' total effect on the target variable was significant. The total effect of exogenous variables including openness, neuroticism, and conscientiousness by the intermediation of deep and surface learning approaches (as The endogenous independent variables' total effect on the target variable was significant. The total effect of endogenous variables including deep and surface intermediary variables), and judicial, executive, and legislative thinking styles on academic achievement were equal to +0.19, -0.16, and +0.15 respectively learning approaches by the intermediation of judicial, executive, and legislative thinking styles on academic achievement were equal to +0.27, and -0.23 respectively. Comparing the direct and total effects of the research variables identifies the important role of the research intermediary variables (learning approaches and thinking styles) in explaining the academic achievement.
A number of indicators of goodness-of-fit have been recommended to test a hypothesized model including Goodness of Fit Index (GFI), Adjusted Goodness of Fit Index (AGFI), Root Mean Square Error of Approximation (RMSEA), and Chi-Square statistic and its associated degrees of freedom (df). Here we used 4 indices for evaluation of model fit and the results (GFI= 0.97), (AGFI=0.95), (RMSE= 0.95), and (Chi-Square=8.45, df=3) indicated that the model provides a good fit to the data.
Model II
We discussed on the direct effect of exogenous variables (personality factors) on endogenous independent variables (learning approaches) above so their results will not be repeated in the discussion of other models. The effects of exogenous variables (personality factors) and independent endogenous variables (deep and surface learning approaches, and levels dimension of thinking styles) on dependent exogenous variable (academic achievement) indicated that the model account for 22% of the variance in academic achievement. Agreeableness, neuroticism, and extroversion had a significant effect on global and local thinking styles. Their regression coefficients (beta) were +0.09, -0.13, and +0.12 respectively.
The endogenous independent variables (deep and surface approaches) had a significant effect on global and local thinking styles (levels dimension). Their regression coefficients (beta) were +0.11, and
-0.09 respectively. Global thinking style (levels dimension) had a significant effect on academic achievement but local thinking style had not a significant effect. Their regression coefficients (beta) were +0.06, and -0.04 respectively.
The exogenous variables' total effect on the target variable was significant. The total effect of exogenous variables including openness, neuroticism, conscientiousness, and extroversion by the intermediation of deep and surface learning approaches (as intermediary variables), and global and local thinking styles on academic achievement were equal to +0.19, -0.17, +0.15 and +0.12 respectively.
The endogenous independent variables' total effect on the target variable was significant. The total effect of endogenous variables including deep and surface learning approaches by the intermediation of global and local thinking styles on academic achievement were equal to +0.26, and -0.22 respectively. Comparing the direct and total effects of the research variables identifies the important role of the research intermediary variables (learning approaches and thinking styles) in explaining the academic achievement.
The indicators of goodness-of-fit (GFI= 0.96), (AGFI=0.94), (RMSE= 0.05), and (Chi-Square=12,
df=5) indicated that the model provides a good fit to the data.
Model III
The effects of exogenous variables (personality factors) and independent endogenous variables (deep and surface learning approaches, and forms dimension of thinking styles) on dependent exogenous variable (academic achievement) indicated that the model account for 15% of the variance in academic achievement. Extroversion had
a significant effect on hierarchical thinking style (forms dimension). Their regression coefficient (beta) was equal to +0.14. Interestingly, other styles which were present in the forms dimension including monarchic style, oligarchic style, and anarchic style were deleted from the model because the direct effect of exogenous variables (personality factors) and learning approaches on them was not significant as well as they had no direct effect on academic achievement.
The endogenous independent variables (deep and surface approaches) had a significant effect on hierarchical thinking styles (forms dimension). Their regression coefficients (beta) were +0.14, and -0.10 respectively. Hierarchical thinking style (forms dimension) had a significant effect on academic achievement and its regression coefficient (beta) was equal to +0.16.
The exogenous variables' total effect on the target variable was significant. The total effect of exogenous variables including openness, neuroticism, conscientiousness, and extroversion by the intermediation of deep and surface learning approaches (as intermediary variables), and hierarchical thinking styles on academic achievement were equal to +0.19, -0.16, +0.15 and
+0.13 respectively.
The endogenous independent variables' total effect on the target variable was significant. The total effect of endogenous variables including deep and surface learning approaches by the intermediation of hierarchical thinking styles on academic achievement were equal to +0.27, and -0.22 respectively. Comparing the direct and total effects of the research variables identifies the important role of the research intermediary variables (learning approaches and thinking styles) in explaining the academic achievement.
The indicators of goodness-of-fit (GFI= 0.98), (AGFI=0.95), (RMSE= 0.04), and (Chi-Square=10,
df=6) indicated that the model provides a good fit to the data.
Model IV
The effects of exogenous variables (personality factors) and independent endogenous variables (deep and surface learning approaches, and scopes dimension of thinking styles) on dependent
exogenous variable (academic achievement) indicated that the model account for 22% of the variance in academic achievement. Openness and extroversion had a significant effect on liberal thinking style (scopes dimension). Their regression coefficients (beta) were equal to +0.16, and +0.13 respectively. Openness and neuroticism had a
significant effect on conservative thinking style.
Their regression coefficients (beta) were equal to -0.12, and +0.18 respectively.
The endogenous independent variables (deep and
surface approaches) had a significant effect on
liberal and conservative thinking styles. Their regression coefficients (beta) were +0.14, and -0.12
(deep approach); and -0.11, and +0.10 (surface
approach) respectively. Liberal and conservative
thinking styles (scopes dimension) had a significant
effect on academic achievement and their regression
coefficients (beta) were equal to +0.18, and -0.14.
The exogenous variables' total effect on the target
variable was significant. The total effect of exogenous variables including openness, neuroticism, conscientiousness, and extroversion by
the intermediation of deep and surface learning
approaches (as intermediary variables), and liberal
and thinking styles on academic achievement were
equal to +0.20, -0.16, +0.14 and +0.13 respectively.
The endogenous independent variables' total effect on the target variable was significant. The
total effect of endogenous variables including deep
and surface learning approaches by the intermediation of liberal and conservative thinking styles on academic achievement were equal to +0.27, and -0.23 respectively. Comparing the direct and total effects of the research variables identifies the important role of the research intermediary variables (learning approaches and thinking styles) in explaining the academic achievement.
The indicators of goodness-of-fit (GFI= 0.99), (AGFI=0.94), (RMSE= 0.03), and (Chi-Square=8, df=4) indicated that the model provides a good fit to the data. 5 Discussion and implications Considerable amount of research have examined the relationships between learning approaches and academic achievement (Biggs, 1998; Watkings and Hotty, 1981, as cited by Zhang 2003; Sadler-Smith & Tsang, 1998). From the Pintrich et al. points of
view (1993) deep processing strategies are thought to lead to greater academic success than surface processing strategies. Our research found that learning approaches in the model which has been provided to explain the academic achievement play different intermediary roles. In other word, our results indicated that deep learning approach play an intermediary role while surface learning approach play an inhibitory role in studying the relationships
between personality factors and academic achievement. Our results also showed that if we control the effect of personality factor the learning approaches may predict the academic achievement independently. Research results indicated that neuroticism has a
negative significant effect on academic achievement. Zhang (2003) emphasized that students with high emotional instability and low
self-esteem (high neuroticism) avoid making
mistakes. Thus they feel more comfort with those tasks that require reproduction of learned materials. These findings are in line with the results of the research that have been conducted by Ruff et al. (2004), Masgeriuo et al. (1997r), Takomoto-Chack (1981 as cited by Diseth, 2003), McCrae (1978 as cited by Diseth, 2003), Blickle, 1996), (Schouwenburg & Kossowska, 1999), and Blickle (1996).
Research results indicated that neuroticism has a negative significant effect on academic achievement. Zhang (2003) emphasized that students with high emotional instability and low self-esteem (high neuroticism) avoid making mistakes. Thus they feel more comfort with those tasks that require reproduction of learned materials. These findings are in line with the results of the research that have been conducted by Ruff et al. (2004), Masgeriuo et al. (1997r), Takomoto-Chack (1981 as cited by Diseth, 2003), McCrae (1978 as cited by Diseth, 2003), Blickle, 1996), Neuroticism showed positive significant correlation with executive and local thinking styles; and conservative thinking styles showed negative significant correlation with legislative, liberal, and hierarchical thinking styles. Persons with high levels of neuroticism feel more comfort from working in highly structured and organized conditions (executive) and prefer that the process of doing tasks to be told and described for them -where the students look for authorities to tell them what to do- (conservative). In other word, persons with emotional instability don't like to work in situations that need new ideas (legislative) and avoid dealing with tasks that require new solutions. People with high score in neuroticism scale are characterized by emotional instability, high irritability, pessimisms, and low self-esteem. Persons with high levels of neuroticism are reactive to stimuli in their environment. They more easily become unstable, worried, temperamental and sad (Howard and Howard, 1998). It seems that such persons feel more comfort from working in highly structured situations and adhere to the available rules when doing their tasks. These characteristics reflect the characteristics of people with executive and conservative styles. The results of present research are in line with the results of Zhang's (2001, 2002) research which emphasized on the relations of neuroticism to executive and conservative thinking styles. Zhang in these research indicated that persons with low self-esteem are tended to using executive and conservative thinking styles. Shadbold (1978, as cited by Zhang 2002) indicated that students with high scores in neuroticism scale prefer structured teaching methods to un-structured teaching methods.
The results also showed that openness has significant positive correlation to judicial, global, and liberal thinking styles; and has significant negative correlation to executive and conservative thinking styles. As mentioned above, persons with high score in openness are imaginative and creative (legislative). These persons tend to do tasks in different ways and don't like to follow affairs in conventional ways (liberal). One of the implied meanings for openness emphasizes that these persons are more willing to interact with other people (external). Also, these people criticize different ideas using differentiation and discrimination ability (judicial).
Finally, conscientiousness is related to a range of thinking styles especially hierarchical thinking style. As discussed before, conscientiousness is characterized by being purposeful, responsible, and
to have a strong will. The conscientious persons strive hard to achieve their goals. They may act creatively (legislative) or analysis and process information (judicial), however, most of the time they prioritize their tasks (hierarchical), and consequently they may use other styles to aid them
in achieving their goals.
Conclusion
The research results which were discussed in the previous section indicated that judicial, hierarchical, legislative, and liberal thinking styles are correlated positively; and executive and conservative thinking styles are correlated negatively with the students' academic achievement. The results of present research regarding the positive relationships between academic achievement and hierarchical, judicial, legislative, and liberal thinking styles are broadly in line with the results of research conducted by Sternberg and Grigorenko (1993, 1997) on the one hand; and its results regarding the negative effect of executive and conservative thinking styles are contradictory with the results of research conducted by Zhang (2001a, 2001b, 2002), Zhang and Sternberg's (1998), Bernard et al. (2002), and Cano-Garcia and Hallgesh (2000).
The results of the present research regarding the effect of thinking styles on academic achievement point out that the thinking styles that are creativity generating and require higher levels of cognitive are
of utmost importance in education field. As it was cited before Zhang (2001) divided thinking styles into two types. People who employed the Type 1 thinking styles (legislative, judicial, global, hierarchical, and liberal styles) tended to be norm-challenging and risk-taking and they were creativity
generating and required complex information processing. People preferred Type 2 thinking styles (executive, local, and conservative styles) required simplistic information processing, they tended to be norm-favoring and authority-oriented. Now, this question needs to be posed that why Type 2 or creativity generating thinking styles can be regarded as the facilitator of students' academic achievement. The answer is that because the study of effect of thinking styles on academic achievement put emphasis on the role of non academic variables on academic achievement. Thus, different studies using the theory of mental self-government indicate the determinant and predictive part that thinking styles could play in academic achievement. Sternberg (1988, 1997 as cited by Zhang 2004) proposed a comprehensive theory of mental self-government which explained the profile of style of an individual using a structure of a government. Thinking styles are somehow social in nature. Success in different subject matters may require different thinking styles thus 'different styles are suitable and effective for various tasks (Sternberg, 1997)'. Studying different subject matters indicate, unexpectedly, that students may have special talents for using specific thinking styles.
Another question need to be asked is that whether we can claim, by the back up of empirical findings and theoretical arguments, that the assessment of thinking styles is fruitless or not. To answer this question we should control the effects of personality factors and then study the direct effect of thinking styles on academic achievement and or study the indirect effect of personality factors on academic achievement. The results of present research indicated that the indirect effect of personality factors on academic achievement by the intermediation of thinking styles is great in comparison to the direct effect of personality factors on academic achievement. Thus, despite the fact that Zhang (2001, 2002) emphasized on an interface between the personality factors and thinking styles, there is no logical reasons that lead us to ignore the role of thinking styles in the study of academic achievement in parallel with personality factors.
The results of present research regarding the effect of learning approaches on thinking styles are in line with the results gained by Zhang (2000) and Zhang and Sternberg (2000). In fact it can be noted that the persons' approach to learning determines how their abilities will be employed both in terms of quality and quantity. Accordingly, learners with deep approach tend to use creativity- generating and complex thinking styles while learners with surface approach tend to use simplistic and norm-favoring thinking styles.
The indirect effect of openness and conscientiousness on academic achievement through deep learning approach was significant and positive, while the indirect effect of openness and conscientiousness on academic achievement through surface learning approach was significant and negative and the indirect effect of neuroticism on academic achievement through surface learning approach was significant and positive. Also, the indirect effect of openness through judicial thinking style, neuroticism through executive thinking style, conscientiousness through legislative thinking style, openness and extraversion through global and liberal thinking styles, neuroticism and openness through conservative thinking style, and extraversion through external thinking style on academic achievement was significant. The total indirect effect of openness, conscientiousness, neuroticism, and extroversion by intermediation of learning approaches and thinking styles dimensions on academic achievement was significant. In sum, the results of present research regarding the indirect effect and total effect of independent exogenous variables on dependent endogenous variable (academic achievement) emphases on the intermediary role of learning approaches and thinking styles variables. Considering the practical implications of the present research are of utmost importance.
References:
[1]Beddoes-Jones, F (2002) Thinking Styles: The design and development of a new psychometric instrument. Thinking Styles Symposium, BJA Associates Ltd. fiona.bj@https://www.doczj.com/doc/d81725490.html,
[2]Bernardo, A.B., Zhang, L.F., & Callueng CM (2002) Thinking styles and academic achievement among Filipino students. Journal of Genet Psychol.
2002; 163(2):149-63
[3]Cano-Garcia, F.,& Hughes, E.H. (2000). earning and thinking styles: an analysis of their interrelationship and in fluence on academic achievement. Education Psychology, 20,413-430.
[4]Duff, Angus, Boyle, Elizabeth A. & Ferguson, R. John (2003) The relationship between personality, approach to learning and academic achievement
[5] Ellis, Aleksander P. J., Hollenbeck, John R., Porter, Christopher O. L. H., Ilgen, Daniel R., West, Bradley J., Moon, Henry (2003), Team Learning: Collectively Connecting the Dots, Journal of Applied Psychology Copyright 2003 by the American Psychological Association, Inc.2003, Vol. 88, No. 5, 821–835
[6] Ferri, Rita Borromeo and Kaiser, Gabriele (2001) Chapter 10: First Results of a Study on Different Mathematical Thinking Styles of School Children.
[7] Furnham A, Jackson CJ and Miller T (1999). Personality learning style and work achievement. Personality and Individual Differences, 27, 1113–1122. [8] Gignac, Gilles E. , Stough, Con, Loukomitis, Sue (2003) Openness, intelligence, and self-report intelligence, Neuropsychology Laboratory, Swinburne Centre for Neuropsychology, Swinburne University of Technology, PO Box 218, Hawthorn 3122, Melbourne, Victoria, Australia
[9] Gross, James J., John, Oliver P. (2003) Individual Differences in Two Emotion Regulation Processes: Implications for Affect, Relationships, and Well-Being, Journal of Personality and Social Psychology, 2003, Vol. 85, No. 2, 348–362
[10] Heinstr?m, Jannica (2000) "The impact of
personality and approaches to learning on information
behaviour." Information Research , 5(3) Available at:
https://www.doczj.com/doc/d81725490.html,/ir/5-3/paper78.html
[11] Heinstr?m, Jannica (2005) Fast surfing, broad
scanning and deep diving The influence of personality
and study approach on students’ information-seeking
behavior, Department of Social and Political
Sciences/Information Studies, Journal of Documentation, Vol. 61 No. 2, 2005, pp. 228-247
[12] Hogan, Michael J. (2006) Against didacticism:
A psychologist’s view, Educational Research and
Reviews Vol. 1 (7), pp. 206-212, October 2006
Available online at http://
https://www.doczj.com/doc/d81725490.html,/ERR, Academic Journals,
Department of Psychology, NUI, Galway, Ireland.
Email: michael.hogan@nuigalway.ie
[13] Huws,Nia, Reddy, Peter & Talcott, Joel (2006)
Predicting university success in psychology: are
subject-specific skills important? Psychology Learning
and Teaching 5 (2), Aston University, UK
[14] Jarvis, Matt (2006) Learning strategies and
achievement in AS-level psychology: an exploratory
study, Totton College
[15] Lizotte, Rick (2005) Scholarship of Teaching and Learning Project: The Effect of Learning Styles in Teaching and Learning
[16] Sternberg, Robert J. (2006) Creating a Vision of Creativity: The First 25 Years, Yale University, Psychology of Aesthetics, Creativity, and the Arts Copyright 2006 by the American Psychological Association, 2006, Vol. S, No. 1, 2–12
[17] Nielsen, Tine, Kreiner, Svend & Styles, Irene (2004) Mental Self-Government: Development of the
Additional Democratic Learning Style using Rasch Measurement Models, Denmark, 18-20 August 2004 [18] Simon Cheng (2006) Thinking Styles in Same-sex Friendship: A Test of Attraction-similarities Hypothesis, The University of Hong Kong
[19] Srivastava, Sanjay, John, Oliver P, Gosling, Samuel D., Potter, Jeff (2003) Development of Personality in Early and Middle Adulthood: Set Like Plaster or Persistent Change? Journal of Personality and Social Psychology, 2003, Vol. 84, No. 5, 1041–1053
[20] Wagerman, Seth A., Funder, David C. (2006) Acquaintance reports of personality and academic achievement: A case for conscientiousness, Department of Psychology, University of California, Riverside, CA 92521, USA, Journal of Research in Personality (2006)
[21] Williams, Tracy L., Turner, Ronna C., Vickers, Ken, & Salamo, Greg (2004) Personality and learning style differences in graduate science programs incorporating business skills trading, Paper presented at the 2004 Annual Meeting of the American Educational Research Association San Diego, CA.
[22] Yunfeng He (2002) The Relationship Between University Students’ Personal Interests and Their Thinking Styles, Faculty of Education, The University of Hong Kong, Pokfulam Road, Hong Kong. OR: Faculty of Law and Politics, Shanghai Teachers University, Shanghai 200234, P. R. China. Email: hyfengb@https://www.doczj.com/doc/d81725490.html,
[23] Zeegers, Peter (2002) A Revision of the Biggs’ Study Process Questionnaire (R-SPQ), Higher Education Research & Development, Vol. 21, No. 1, 2002, Flinders University [24] Zhang (2000a)"Are Learning Approaches and
Thinking Styles Related? A Study in Two Chinese Populations". Journal of Psychology 134(5):469-89.
[25] Zhang (2000b)"Are Thinking Styles and
Personality Types Related?" Educational Psychology
20(3):271-83.
[26] Zhang L.F. & Sternberg RJ. (2002) Are learning
approaches and thinking styles related? A study in two
Chinese populations. Journal of Psychology 2000 Sep;
134(5):469-89. Department of Education, the
University of Hong Kong. lfzhang@hkucc.hku.hk
[27] Zhang L.F. (2002). Thinking styles and the big
five personality traits. Educational psychology, 22(1),
17-31.
信号与系统》复习参考练习题一、单项选择题：
14、已知连续时间信号,) 2(100) 2(50sin )(--= t t t f 则信号t t f 410cos ·)(所占有的频带宽度为（） A ．400rad ／s B 。200 rad ／s C 。100 rad ／s D 。50 rad ／s
f如下图（a）所示，其反转右移的信号f1(t) 是（） 15、已知信号)(t f如下图所示，其表达式是（） 16、已知信号)(1t A、ε（t）＋2ε(t－2)－ε(t－3) B、ε(t－1)＋ε(t－2)－2ε(t－3) C、ε(t)＋ε(t－2)－ε(t－3) D、ε(t－1)＋ε(t－2)－ε(t－3) 17、如图所示：f（t）为原始信号，f1(t)为变换信号，则f1(t)的表达式是（） A、f(－t+1) B、f(t+1) C、f(－2t+1) D、f(－t/2+1) 18、若系统的冲激响应为h(t),输入信号为f(t),系统的零状态响应是（）
19。信号)2(4 sin 3)2(4 cos 2)(++-=t t t f π π 与冲激函数)2(-t δ之积为（ ） A 、2 B 、2)2(-t δ C 、3)2(-t δ D 、5)2(-t δ ，则该系统是（）＞－系统的系统函数．已知2]Re[,6 51 )(LTI 202s s s s s H +++= A 、因果不稳定系统 B 、非因果稳定系统 C 、因果稳定系统 D 、非因果不稳定系统 21、线性时不变系统的冲激响应曲线如图所示，该系统微分方程的特征根是（ ） A 、常数 B 、 实数 C 、复数 D 、实数+复数 22、线性时不变系统零状态响应曲线如图所示，则系统的输入应当是（ ） A 、阶跃信号 B 、正弦信号 C 、冲激信号 D 、斜升信号 23. 积分 ?∞ ∞ -dt t t f )()(δ的结果为( ) A )0(f B )(t f C.)()(t t f δ D.)()0(t f δ 24. 卷积)()()(t t f t δδ**的结果为( ) A.)(t δ B.)2(t δ C. )(t f D.)2(t f
信号与系统期末考试试题 一、选择题（共10题，每题3分 ，共30分，每题给出四个答案，其中只有一个正确的） 1、 卷积f 1(k+5)*f 2(k-3) 等于 。 （A ）f 1(k)*f 2(k) （B ）f 1(k)*f 2(k-8)（C ）f 1(k)*f 2(k+8)（D ）f 1(k+3)*f 2(k-3) 2、 积分 dt t t ? ∞ ∞ --+)21()2(δ等于 。 （A ）1.25（B ）2.5（C ）3（D ）5 3、 序列f(k)=-u(-k)的z 变换等于 。 （A ） 1-z z （B ）-1-z z （C ）11-z （D ）1 1--z 4、 若y(t)=f(t)*h(t),则f(2t)*h(2t)等于 。 （A ） )2(41t y （B ）)2(21t y （C ）)4(41t y （D ）)4(2 1 t y 5、 已知一个线性时不变系统的阶跃相应g(t)=2e -2t u(t)+)(t δ,当输入f(t)=3e —t u(t)时，系 统的零状态响应y f (t)等于 （A ）(-9e -t +12e -2t )u(t) （B ）(3-9e -t +12e -2t )u(t) （C ）)(t δ+(-6e -t +8e -2t )u(t) （D ）3)(t δ +(-9e -t +12e -2t )u(t) 6、 连续周期信号的频谱具有 （A ） 连续性、周期性 （B ）连续性、收敛性 （C ）离散性、周期性 （D ）离散性、收敛性 7、 周期序列2)455.1(0 +k COS π的 周期N 等于 （A ） 1（B ）2（C ）3（D ）4 8、序列和 ()∑∞ -∞ =-k k 1δ等于 （A ）1 (B) ∞ (C) ()1-k u (D) ()1-k ku 9、单边拉普拉斯变换()s e s s s F 22 12-+= 的愿函数等于 ()()t tu A ()()2-t tu B ()()()t u t C 2- ()()()22--t u t D 10、信号()()23-=-t u te t f t 的单边拉氏变换()s F 等于 ()A ()()()232372+++-s e s s ()() 2 23+-s e B s
M A T L A B模型预测控制工具箱函数 8.2系统模型建立与转换函数 前面读者论坛了利用系统输入/输出数据进行系统模型辨识的有关函数及使用方法，为时行模型预测控制器的设计，需要对系统模型进行进一步的处理和转换。MATLAB的模型预测控制工具箱中提供了一系列函数完成多种模型转换和复杂系统模型的建立功能。 在模型预测控制工具箱中使用了两种专用的系统模型格式，即MPC状态空间模型和MPC传递函数模型。这两种模型格式分别是状态空间模型和传递函数模型在模型预测控制工具箱中的特殊表达形式。这种模型格式化可以同时支持连续和离散系统模型的表达，在MPC传递函数模型中还增加了对纯时延的支持。表8-2列出了模型预测控制工具箱的模型建立与转换函数。 表8-2模型建立与转换函数 8.2.1模型转换 在MATLAB模型预测工具箱中支持多种系统模型格式。这些模型格式包括： ①通用状态空间模型； ②通用传递函数模型； ③MPC阶跃响应模型； ④MPC状态空间模型；
⑤MPC传递函数模型。 在上述5种模型格式中，前两种模型格式是MATLAB通用的模型格式，在其他控制类工具箱中，如控制系统工具箱、鲁棒控制工具等都予以支持；而后三种模型格式化则是模型预测控制工具箱特有的。其中，MPC状态空间模型和MPC传递函数模型是通用的状态空间模型和传递函数模型在模型预测控制工具箱中采用的增广格式。模型预测控制工具箱提供了若干函数，用于完成上述模型格式间的转换功能。下面对这些函数的用法加以介绍。 1．通用状态空间模型与MPC状态空间模型之间的转换 MPC状态空间模型在通用状态空间模型的基础上增加了对系统输入/输出扰动 和采样周期的描述信息，函数ss2mod()和mod2ss()用于实现这两种模型格式之间的转换。 1）通用状态空间模型转换为MPC状态空间模型函数ss2mod() 该函数的调用格式为 pmod=ss2mod(A,B,C,D) pmod=ss2mod(A,B,C,D,minfo) pmod=ss2mod(A,B,C,D,minfo,x0,u0,y0,f0) 式中，A,B,C,D为通用状态空间矩阵； minfo为构成MPC状态空间模型的其他描述信息，为7个元素的向量，各元素分别定义为： ◆minfo(1)=dt，系统采样周期，默认值为1； ◆minfo(2)=n，系统阶次，默认值为系统矩阵A的阶次； ◆minfo(3)=nu，受控输入的个数，默认值为系统输入的维数； ◆minfo(4)=nd，测量扰的数目，默认值为0； ◆minfo(5)=nw，未测量扰动的数目，默认值为0； ◆minfo(6)=nym，测量输出的数目，默认值系统输出的维数； ◆minfo(7)=nyu，未测量输出的数目，默认值为0； 注：如果在输入参数中没有指定m i n f o，则取默认值。 x0,u0,y0,f0为线性化条件，默认值均为0； pmod为系统的MPC状态空间模型格式。 例8-5将如下以传递函数表示的系统模型转换为MPC状态空间模型。 解：MATLAB命令如下：
信科0801《信号与系统》复习参考练习题 参考答案 信号与系统综合复习资料 考试方式：闭卷 考试题型：1、简答题（5个小题），占30分；计算题（7个大题），占70分。 一、简答题： 1．dt t df t f x e t y t ) ()()0()(+=-其中x(0)是初始状态， 为全响应，为激励，)()(t y t f 试回答该系统是否是线性的？[答案：非线性] 2．)()(sin )('t f t ty t y =+试判断该微分方程表示的系统是线性的还是非线性的， 是时变的还是非时变的？[答案：线性时变的] 3．已知有限频带信号)(t f 的最高频率为100Hz ，若对)3(*)2(t f t f 进行时域取样， 求最小取样频率s f =？[答案：400s f Hz =] 4．简述无失真传输的理想条件。[答案：系统的幅频特性为一常数，而相频特性为通过原点的直线] 5．求[]?∞ ∞ --+dt t t e t )()('2δδ的值。[答案：3] 6．已知)()(ωj F t f ?，求信号)52(-t f 的傅立叶变换。 [答案：521(25)()22 j f t e F j ωω --?]
7．已知)(t f 的波形图如图所示，画出)2()2(t t f --ε的波形。 [答案： ] 8．已知线性时不变系统，当输入)()()(3t e e t x t t ε--+=时，其零状态响应为 )()22()(4t e e t y t t ε--+=，求系统的频率响应。[答案： ()) 4)(2(52)3(++++ωωωωj j j j ] 9．求象函数2 ) 1(3 2)(++= s s s F ,的初值)0(+f 和终值)(∞f 。 [答案：)0(+f =2，0)(=∞f ] 10．若LTI 离散系统的阶跃响应为)(k g ，求其单位序列响应。 其中：)()2 1 ()(k k g k ε=。 [答案：1111 ()()(1)()()()(1)()()(1)222 k k k h k g k g k k k k k εεδε-=--=--=--] 11．已知()1 1 , 0,1,20 , k f k else ==??? ，()2 1 , 0,1,2,3 0 , k k f k else -==??? 设()()()12f k f k f k =*，求()3?f =。[答案：3] 12．描述某离散系统的差分方程为()()()122()y k y k y k f k +---=
神经网络模型预测控制器 摘要：本文将神经网络控制器应用于受限非线性系统的优化模型预测控制中，控制规则用一个神经网络函数逼近器来表示，该网络是通过最小化一个与控制相关的代价函数来训练的。本文提出的方法可以用于构造任意结构的控制器，如减速优化控制器和分散控制器。 关键字：模型预测控制、神经网络、非线性控制 1.介绍 由于非线性控制问题的复杂性，通常用逼近方法来获得近似解。在本文中，提出了一种广泛应用的方法即模型预测控制（MPC），这可用于解决在线优化问题，另一种方法是函数逼近器，如人工神经网络，这可用于离线的优化控制规则。 在模型预测控制中，控制信号取决于在每个采样时刻时的想要在线最小化的代价函数，它已经广泛地应用于受限的多变量系统和非线性过程等工业控制中[3,11,22]。MPC方法一个潜在的弱点是优化问题必须能严格地按要求推算，尤其是在非线性系统中。模型预测控制已经广泛地应用于线性MPC问题中[5]，但为了减小在线计算时的计算量，该部分的计算为离线。一个非常强大的函数逼近器为神经网络，它能很好地用于表示非线性模型或控制器，如文献[4,13,14]。基于模型跟踪控制的方法已经普遍地应用在神经网络控制，这种方法的一个局限性是它不适合于不稳定地逆系统，基此本文研究了基于优化控制技术的方法。 许多基于神经网络的方法已经提出了应用在优化控制问题方面，该优化控制的目标是最小化一个与控制相关的代价函数。一个方法是用一个神经网络来逼近与优化控制问题相关联的动态程式方程的解[6]。一个更直接地方法是模仿MPC方法，用通过最小化预测代价函数来训练神经网络控制器。为了达到精确的MPC技术，用神经网络来逼近模型预测控制策略，且通过离线计算[1,7.9,19]。用一个交替且更直接的方法即直接最小化代价函数训练网络控制器代替通过训练一个神经网络来逼近一个优化模型预测控制策略。这种方法目前已有许多版本，Parisini[20]和Zoppoli[24]等人研究了随机优化控制问题，其中控制器作为神经网络逼近器的输入输出的一个函数。Seong和Widrow[23]研究了一个初始状态为随机分配的优化控制问题，控制器为反馈状态，用一个神经网络来表示。在以上的研究中，应用了一个随机逼近器算法来训练网络。Al-dajani[2]和Nayeri等人[15]提出了一种相似的方法，即用最速下降法来训练神经网络控制器。 在许多应用中，设计一个控制器都涉及到一个特殊的结构。对于复杂的系统如减速控制器或分散控制系统，都需要许多输入与输出。在模型预测控制中，模型是用于预测系统未来的运动轨迹，优化控制信号是系统模型的系统的函数。因此，模型预测控制不能用于定结构控制问题。不同的是，基于神经网络函数逼近器的控制器可以应用于优化定结构控制问题。 在本文中，主要研究的是应用于非线性优化控制问题的结构受限的MPC类型[20,2,24,23,15]。控制规则用神经网络逼近器表示，最小化一个与控制相关的代价函数来离线训练神经网络。通过将神经网络控制的输入适当特殊化来完成优化低阶控制器的设计，分散和其它定结构神经网络控制器是通过对网络结构加入合适的限制构成的。通过一个数据例子来评价神经网络控制器的性能并与优化模型预测控制器进行比较。 2.问题表述 考虑一个离散非线性控制系统： 其中为控制器的输出，为输入，为状态矢量。控制
信科0801《信号与系统》复习参考练习题一、单项选择题：
14、已知连续时间信号,) 2(100)2(50sin )(--=t t t f 则信号t t f 410cos ·)(所占有的频带宽度为（） A ．400rad ／s B 。200 rad ／s C 。100 rad ／s D 。50 rad ／s
f如下图（a）所示，其反转右移的信号f1(t) 是（） 15、已知信号)(t f如下图所示，其表达式是（） 16、已知信号)(1t A、ε（t）＋2ε(t－2)－ε(t－3) B、ε(t－1)＋ε(t－2)－2ε(t－3) C、ε(t)＋ε(t－2)－ε(t－3) D、ε(t－1)＋ε(t－2)－ε(t－3) 17、如图所示：f（t）为原始信号，f1(t)为变换信号，则f1(t)的表达式是（） A、f(－t+1) B、f(t+1) C、f(－2t+1) D、f(－t/2+1)
18、若系统的冲激响应为h(t),输入信号为f(t),系统的零状态响应是（ ） 19。信号)2(4sin 3)2(4cos 2)(++-=t t t f π π 与冲激函数)2(-t δ之积为（ ） A 、2 B 、2)2(-t δ C 、3)2(-t δ D 、5)2(-t δ ，则该系统是（）＞－系统的系统函数．已知2]Re[,6 51)(LTI 202s s s s s H +++= A 、因果不稳定系统 B 、非因果稳定系统 C 、因果稳定系统 D 、非因果不稳定系统 21、线性时不变系统的冲激响应曲线如图所示，该系统微分方程的特征根是（ ） A 、常数 B 、 实数 C 、复数 D 、实数+复数 22、线性时不变系统零状态响应曲线如图所示，则系统的输入应当是（ ） A 、阶跃信号 B 、正弦信号 C 、冲激信号 D 、斜升信号
基于logistic模型的2014年影响中国各省城市化水平的 经济地理因素分析 摘要：本文利用2013年中国31个省份的数据，从经济与地理位置两个因素出发，运用logistic回归的方法在SPSS软件上进行分析。结果显示：中国城市化发展水平不仅与经济密切相关，而且与其地理位置也有很大的关系，地区间城市化发展水平差距较明显，城市化各方面的因素水平发展不平衡。 关键词：logistic模型，城市化水平，SPSS软件
目录 一、引言 (3) 二、Logistic模型 (3) 1. 基本概念 (3) 2. 统计原理 (4) （1）logit变换 (4) （2）Logistic回归模型 (4) （3）统计检验 (4) 三、基于logistic模型的我国各省城市化水平影响因素实证分析 (5) 1．数据来源与说明 (5) 2．模型检验 (5) 3．模型的建立与预测 (7) 四、结论 (7) 参考文献 (8)
一、引言 城市化的定义众多，本文参照《中华人民国国家标准城市规划术语》，认为城市化是“人类生产与生活方式由农村型向城市型转化的历史过程，主要表现为农村人口转化为城市人口及城市不断发展完善的过程。”城市化是一个系统的动态过程，包含了人口、经济、社会、城市建设等各方面变化的影响。它是经济发展和社会进步的必然结果，反过来也推动了经济的发展和社会的进步。 中国大陆的城市化进程在不同的时期具有不同的特点，总的来看城市化水平普遍较低，并已成为制约国家经济、社会和谐发展的主要原因之一。因而，各地区普遍把推进城市化进程作为经济、社会发展战略的一项重要目标选择。当前中国大陆已经进入了城市化水平的持续上升发展时期，此时对这样一个过程实施有效、客观、科学、动态的监测，从而及时发现并解决城市化进程中出现的难题，就必须加强对中国大陆城市化水平质与量等方面的考察和研究。这对于我们这样一个人口众多、区域经济发展不平衡的国家尤为重要。 本文不仅分析影响城市化水平的经济因素，还加入了地理位置对其城市化发展的影响。由于地理因素数据不是数值型变量，因此我们引用logistic回归方法对其进行建模。 二、Logistic模型 1.基本概念 Logistic回归分析就是针对因变量是定型变量的回归分析，这与一般的回归分析不同。在实际生活中，我们会经常遇到因变量是定型
一．填空题(本大题共10空，每空2分，共20分。) 1．()*(2)k k εδ-= . 2． sin()()2 t d π τδττ-∞ + =? . 3. 已知信号的拉普拉斯变换为 1 s a -，若实数a ，则信号的傅里叶变换不存在. 4. ()()()t h t f t y *=，则()=t y 2 . 5. 根据Parseval 能量守恒定律，计算?∞ ∞-=dt t t 2 )sin ( . 6. 若)(t f 最高角频率为m ω，则对 )2()4()(t f t f t y =取样，其频谱不混迭的最大间隔是 . 7. 某因果线性非时变（LTI ）系统，输入)()(t t f ε=时，输出为： )1()()(t t e t y t --+=-εε；则) 2()1()(---=t t t f εε时，输出)(t y f ＝ . 8. 已知某因果连续LTI 系统)(s H 全部极点均位于s 左半平面，则 ∞→t t h )(的值为 . 9. 若)()(ωj F t f ?，已知)2cos()(ωω=j F ，试求信号)(t f 为 . 10.已知某离散信号的单边z 变换为) 3(,)3)(2(2)(2>+-+=z z z z z z F ，试求其反变换 )(k f = . 二．选择题(本大题共5小题，每题4分，共20分。) 1.下列信号的分类方法不正确的是 ： A 、数字信号和离散信号 B 、确定信号和随机信号 C 、周期信号和非周期信号 D 、因果信号与反因果信号 2. )]2()()[2()]()2([2)(1--++-+=t t t t t t f εεεε，则)] 1()2 1()[21()(--+-=t t t f t f εε
云南大学信息学院学生实验报告 课程名称：现代控制理论 实验题目：预测控制 小组成员：李博（12018000748） 金蒋彪（12018000747） 专业：2018级检测技术与自动化专业
1、实验目的 (3) 2、实验原理 (3) 2.1、预测控制特点 (3) 2.2、预测控制模型 (4) 2.3、在线滚动优化 (5) 2.4、反馈校正 (5) 2.5、预测控制分类 (6) 2.6、动态矩阵控制 (7) 3、MATLAB仿真实现 (9) 3.1、对比预测控制与PID控制效果 (9) 3.2、P的变化对控制效果的影响 (12) 3.3、M的变化对控制效果的影响 (13) 3.4、模型失配与未失配时的控制效果对比 (14) 4、总结 (15) 5、附录 (16) 5.1、预测控制与PID控制对比仿真代码 (16) 5.1.1、预测控制代码 (16) 5.1.2、PID控制代码 (17) 5.2、不同P值对比控制效果代码 (19) 5.3、不同M值对比控制效果代码 (20) 5.4、模型失配与未失配对比代码 (20)
1、实验目的 （1）、通过对预测控制原理的学习，掌握预测控制的知识点。 （2）、通过对动态矩阵控制（DMC）的MATLAB仿真，发现其对直接处理具有纯滞后、大惯性的对象，有良好的跟踪性和较强的鲁棒性，输入已 知的控制模型，通过对参数的选择，来获得较好的控制效果。 （3）、了解matlab编程。 2、实验原理 模型预测控制(Model Predictive Control，MPC)是20世纪70年代提出的一种计算机控制算法，最早应用于工业过程控制领域。预测控制的优点是对数学模型要求不高，能直接处理具有纯滞后的过程，具有良好的跟踪性能和较强的抗干扰能力，对模型误差具有较强的鲁棒性。因此，预测控制目前已在多个行业得以应用，如炼油、石化、造纸、冶金、汽车制造、航空和食品加工等，尤其是在复杂工业过程中得到了广泛的应用。在分类上，模型预测控制(MPC)属于先进过程控制，其基本出发点与传统PID控制不同。传统PID控制，是根据过程当前的和过去的输出测量值与设定值之间的偏差来确定当前的控制输入，以达到所要求的性能指标。而预测控制不但利用当前时刻的和过去时刻的偏差值，而且还利用预测模型来预估过程未来的偏差值，以滚动优化确定当前的最优输入策略。因此，从基本思想看，预测控制优于PID控制。 2.1、预测控制特点 首先，对于复杂的工业对象。由于辨识其最小化模型要花费很大的代价，往往给基于传递函数或状态方程的控制算法带来困难，多变量高维度复杂系统难以建立精确的数学模型工业过程的结构、参数以及环境具有不确定性、时变性、非线性、强耦合，最优控制难以实现。而预测控制所需要的模型只强调其预测功能，不苛求其结构形式，从而为系统建模带来了方便。在许多场合下，只需测定对象的阶跃或脉冲响应，便可直接得到预测模型，而不必进一步导出其传递函数或状
南湖学院机电系《信号与系统》课程考试试题 2013—2014学年 第 二 学期 N 电信12班级 时量：120分钟 总分：100分 考试形式： 开卷(A) 一、 填空题 （每小题2分，共20分） 1、)2()()(-t t u t f δ=（ ）。 2、=-*-)()(21t t t t f δ（ ）。 3、拉普拉斯变换是把时域信号变换到（ ）。 4、对一个频带限制在0~4KHz 的语音信号进行抽样，则奈奎斯特速率是（ ）。 5、从信号频谱的连续性和离散性来观察，非周期信号的频谱是（ ）的。 6、线性时不变连续因果系统是稳定系统的充分必要条件是)(s H 的极点位于（ ）。 7、信号不失真传输的条件是系统函数=)(ωj H （ ）。 8、若自由响应对应系统微分方程的齐次解，则强迫响应对应系统微分方程的（ ）。 9、零输入线性是指当激励为0时，系统的零输入响应对各（ ）呈线性。 10、采用（ ）滤波器即可从已抽样信号中恢复原模拟信号。 二、选择题 （每小题2分，共20分） 1、信号 x (-n +2) 表示（ ）。 A 、信号x (n )的右移序2 B 、信号x (n )的左移序2 C 、信号x (n )反转再右移序2 D 、信号x (n )反转再左移序2 2、二阶前向差分)(2n x ?的表示式是（ ）。 A 、)()1(2)2(n x n x n x ++++ B 、)()1(2)2(n x n x n x ++-+ C 、)2()1(2)(-+-+n x n x n x D 、)2()1(2)(-+--n x n x n x 3、在以下关于冲击信号)(t δ的性质表达式中，不正确的是 ( )。 A 、? ∞ ∞ -=')()(t dt t δδ B 、?∞ ∞ -='0)(dt t δ C 、 ? ∞ -=t t u dt t )()(δ D 、)()(t t δδ=- 4、下列4个常用信号的傅立叶变换式中，不正确的是( )。 A 、)(21ωπδ? B 、)(200ωωπδω-?t j e C 、()()[]000cos ωωδωωδπω++-?t D 、()()[]000sin ωωδωωδπω++-?j t 5、系统仿真图如图所示，则系统的单位冲激响应)(t h 满足的方程式是（ ）。
【摘要】logistic阻滞增长模型在人口预测中有着广泛应用，应用spss软件能较为简便地进行logistic曲线的拟合。文章介绍了spss拟合logistic人口预测方程的两种方法及其步骤，并通过其结果分析比较二者的优缺点。 【关键词】logistic；spss软件；拟合方法 logistic模型为荷兰数学家及生物学家verhulst.pearl在修正非密度方程时提出，其目的为研究受到生存资源制约的情况下生物种群的增长规律。在logistic模型中，有限空间内种群不能无限增长，而是存在着数量上限。由于自然资源、环境条件等因素对种群的增长起着阻滞作用，并且随着种群数量的增大，阻滞作用逐步增大，即实测增长率是一个减函数，且随着种群数量的增大而减小，当种群数量趋于上限时，种群增长亦趋于稳定。由于logistic 阻滞增长模型所需的数据少，计算简单，对中短期时间内的种群数量预测较为准确，亦常应用于人口预测方面。 一、logistic阻滞增长模型 如上文述，人口增长率为以人口数量x为自变量的函数r（x），这里r（x）为减函数。假设r（x）= r ?sx，s>0，这里r为初始值r（），即当人口无生存环境和资源限制时的固有增长率。当人口数量达到人口最大容量，将有r（）=0，此时人口达到稳定状态。由线性关系r（）=r-s，可得s=r/。假设x是时间t的函数x（t），从而有解变量可分离方程。 二、spss软件拟合logistic人口阻滞增长模型 通过模型方程（ⅰ）可知，logistic模型拟合的重点为参数和的确定。下采用两种spss 软件的回归拟合方法，利用1990-2010年人口调查数据（如表1）进行人口数量的预测。 （一）非线性回归（nonlinear regression）拟合 在spss（spss19.0）的变量视图中定义两变量人口数量x及年份t，在数据视图中由上而下录入人口数据（如图1所示）。 在菜单栏依次选择分析（analyze）―回归（regression）―非线性估计（nonlinear），打开非线性回归窗口。将年末总人口[x]送入因变量一栏，在模型表达式输入框中输入模型公式 a/（1 +（a / 114333 - 1）* exp（- r *（t - 1990）））（如图2）。此处以a代替人口最大容量，由于时间以1990年为初始年份，原方程中的t转为t-1990。选择“参数”项进行参数a和r初始值的设定（如图3），这里a初始值选择人数中的最大值134091（万人），r 的初始值选择1991年的人口增长率0.013，“使用上一分析的起始值”一栏选中，单击“继续”。单击“保存”项，打开对话框如图4，选中预测值和残差项，便于检验模型方程的拟合效果，选择“继续”返回非线性回归窗口，选择“确定”运行。在输出（output）窗口中，可以得到参数a的迭代计算过程、参数估计等内容。由参数估计得参数估计值，=0.0675。r2=1.000。 （二）曲线估计法 采用spss的曲线估计进行模型拟合，须先求参数。对估计的方法很多，这里采用三点法进行求取。 选择分析（analyze）―回归（regression）―曲线估计（curve estimation），打开曲线估计窗口，将年末总人口[x]和年份[t]分别送入因变量和自变量输入框，在“模型”区选中logistic，在上限一栏填入142515.5576，在“保存”对话框中选中预测值和残差，其他依照默认选择。选择“确定”。 三、对两种方法所得拟合方程的讨论 从可决系数r2来看，两种方法所得拟合方程的r2均得1，则两种方法对logistic人口预测模型的拟合性都很好。分别用两种方法所得方程对2011年和2012年的年末人口数进行
智慧树知到《分子生物学》章节测试答案 第一章 1、目前生物遗传信息传递规律中还没有实验证据的是（）. A:A. DNA→RNA B:B. RNA→蛋白质 C:C. RNA→DNA D:D. 蛋白质→DNA 正确答案： D. 蛋白质→DNA 2、从小鼠的一种有夹膜的致病性肺炎球菌中提取出的DNA，可使另一种无荚膜、不具有致病性的肺炎球菌转变为有夹膜并具有致病性的肺炎球菌，而蛋白质、RNA无此作用，由此可以证明（）。 A:DNA是遗传物质，蛋白质是遗传信息的体现者。 B:蛋白质是遗传物质，DNA是遗传信息的体现者。 C:DNA和蛋白质均是遗传物质。 D:RNA是遗传物质，DNA和蛋白质是遗传信息的体现者。 正确答案：DNA是遗传物质，蛋白质是遗传信息的体现者。 3、自然界中以DNA为遗传物质的大多数生物DNA的复制方式为（）。 A:环式 B:半保留 C:D-环式 D:全保留 正确答案：半保留 4、1997年诺贝尔生理医学奖授予美国加州旧金山大学Stanley B.Prusien，表彰他发现朊病毒及其致病机理，请问朊病毒是一种（）？
A:DNA B:葡萄糖 C:蛋白质 D:RNA 正确答案：蛋白质 5、证明DNA复制为半保留复制的科学家为（）。 A:Meselson&stahl B:F. Miesher C:o. Avery D:Chargaff 正确答案：Meselson&stahl 第二章 1、证明DNA是遗传物质的两个关键性实验是：肺炎球菌在老鼠体内的毒性和T2噬菌体感染大肠杆菌。这两个实验中主要的论点证据是（）。 A:从被感染的生物体内重新分离得到DNA作为疾病的致病剂 B:DNA突变导致毒性丧失 C:生物体吸收的外源DNA（而并非蛋白质）改变了其遗传潜能 D:DNA是不能在生物体间转移的，因此它一定是一种非常保守的分子 正确答案：生物体吸收的外源DNA（而并非蛋白质）改变了其遗传潜能 2、物种的C值与其进化复杂性之间无严格对应关系。 A:对 B:错
信号与系统 一、填空题：（30分，每小题3分） 1. 已知f(t)的傅里叶变换为F(jω), 则f(2t-3)的傅里叶变换为。 2、。 3 = 。 4. 已知，则 ; 。 5. 已知，则。 6. 已知周期信号，其基波频率为 rad/s； 周期为 s。 7. 已知，其Z变换 ；收敛域为。 8. 已知连续系统函数，试判断系统的稳定性：。 9．已知离散系统函数，试判断系统的稳定性：。 10．如图所示是离散系统的Z域框图，该系统的系统函数H(z)= 。二．(15分)如下方程和非零起始条件表示的连续时间因果LTI系统， 已知输入时，试用拉普拉斯变换的方法求系统的零状态响应 和零输入响应，以及系统的全响应。 三．（14分） 1 已知,,试求其拉氏逆变换f(t)； 2 已知，试求其逆Z变换。 四（10分）计算下列卷积： 1. ；
2．。 五．（16分）已知系统的差分方程和初始条件为： ， 1. 求系统的全响应y(n)； 2. 求系统函数H(z)，并画出其模拟框图； 六．（15分）如图所示图（a）的系统，带通滤波器的频率响应如图(b)所示，其相位特性，若输入信号为: 试求其输出信号y(t)，并画出y(t)的频谱图。
参考答案一填空题（30分，每小题3分） 2. 1 ； 2. e-2 ; 3. ; 4. 1 ，0 ; ； 6. 2 л ; ，|z|>0； 8. 不稳定； 9. 稳定 10. 二．（15分） 方程两边取拉氏变换： 三．1．（7分）
2．（7分） 四． 1. （5分） 2.（5分） 五．解：（16分） （1）对原方程两边同时Z变换有：（2） 六（15分）
二分类Logistic 回归模型 在对资料进行统计分析时常遇到反应变量为分类变量的资料，那么，能否用类似于线性回归的模型来对这种资料进行分析呢？答案是肯定的。本章将向大家介绍对二分类因变量进行回归建模的Logistic 回归模型。 第一节 模型简介 一、模型入门 在很多场合下都能碰到反应变量为二分类的资料，如考察公司中总裁级的领导层中是否有女性职员、某一天是否下雨、某病患者结局是否痊愈、调查对象是否为某商品的潜在消费者等。对于分类资料的分析，相信大家并不陌生，当要考察的影响因素较少，且也为分类变量时，分析者常用列联表(contingency T able)的形式对这种资料进行整理，并使用2 χ检验来进行分析，汉存在分类的混杂因素时，还可应用Mantel-Haenszel 2χ检验进行统计学检验，这种方法可以很好地控制混杂因素的影响。但是这种经典分析方法也存在局限性，首先，它虽然可以控制若干个因素的作用，但无法描述其作用大小及方向，更不能考察各因素间是否存在交互任用；其次，该方法对样本含量的要求较大，当控制的分层因素较多时，单元格被划分的越来越细，列联表的格子中频数可能很小甚至为0，将导致检验结果的不可靠。最后，2χ检验无法对连续性自变量的影响进行分析，而这将大大限制其应用范围，无疑是其致使的缺陷。 那么，能否建立类似于线性回归的模型，对这种数据加以分析？以最简单的二分类因变量为例来加以探讨，为了讨论方便，常定义出现阳性结果时反应变量取值为1，反之则取值为0 。例如当领导层有女性职员、下雨、痊愈时反应变量1y =，而没有女性职员、未下雨、未痊愈时反应变量0y =。记出现阳性结果的频率为反应变量(1)P y =。 首先，回顾一下标准的线性回归模型：
分子生物学笔记 第一章基因的结构 第一节基因和基因组 一、基因(gene) 是合成一种功能蛋白或RNA分子所必须的全部DNA序列． 一个典型的真核基因包括 ①编码序列—外显子(exon) ②插入外显子之间的非编码序列—内合子(intron) ③5'-端和3'-端非翻译区(UTR) ④调控序列(可位于上述三种序列中) 绝大多数真核基因是断裂基因(split-gene)，外显子不连续。 二、基因组(genome) 一特定生物体的整套(单倍体)遗传物质的总和，基因组的大小用全部DNA的碱基对总数表示。人基因组3X1 09(30亿bp)，共编码约10万个基因。 每种真核生物的单倍体基因组中的全部DNA量称为C值，与进化的复杂性并不一致(C-value Paradox)。 人类基因组计划（human genome project, HGP） 基因组学（genomics）,结构基因组学（structural genomics）和功能基因组学（functional genomics）。 蛋白质组（proteome）和蛋白质组学（proteomics） 第二节真核生物基因组 一、真核生物基因组的特点：， ①真核基因组DNA在细胞核内处于以核小体为基本单位的染色体结构中． ②真核基因组中，编码序列只占整个基因组的很小部分(2—3％)， 三、基因家族(gene family) 一组功能相似且核苷酸序列具有同源性的基因．可能由某一共同祖先基因(ancestral gene)经重复(duplication)和突变产生。 基因家族的特点： ①基因家族的成员可以串联排列在一起，形成基因簇(gene cluster)或串联重复基因(tandemly repeated genes)，如rRNA、tRNA和组蛋白的基因；②有些基因家族的成员也可位于不同的染色体上，如珠蛋白基因；③有些成员不产生有功能的基因产物，这种基因称为假基因(Pseudogene)．Ψa1表示与a1相似的假基因． 四、超基因家族(Supergene family ，Superfamily) 由基因家族和单基因组成的大基因家族，结构上有程度不等的同源性，但功能不同． 第四节细菌和病毒基因组 一、细菌基因组的特点。 1．功能相关的几个结构基因往往串联在—起，受它们上游的共同调控区控制，形成操纵子结构， 2．结构基因中没有内含子，也无重叠现象。 3．细菌DNA大部分为编码序列。 二、病毒基因组的特点 1．每种病毒只有一种核酸，或者DNA，或者RNA； 2．病毒核酸大小差别很大，3X103一3X106bp； 3．除逆病毒外，所有病毒基因都是单拷贝的。 4．大部份病毒核酸是由一条双链或单链分子(RNA或DNA)，仅少数RNA病毒由几个核酸片段组成． 5．真核病毒基因有内含子，而噬菌体（感染细菌的病毒）基因中无内含子． 6．有重叠基因． 第五节染色质和染色体 (二)组蛋白(histone)：一类小的带有丰富正电荷<富含Lys，Arg)的核蛋白，与DNA有高亲和力． (二)．端粒(telomere)：真核生物线状染色体分子末端的DNA区域 端粒DNA的特点： 1、由富含G的简单串联重复序列组成(长达数kb)． 人的端粒DNA重复序列：TTAGGC。
《信号与系统》期末试卷A 卷 班级： 学号：__________ 姓名：________ _ 成绩：_____________ 一． 选择题（共10题，20分） 1、n j n j e e n x )3 4( )3 2(][ππ+=，该序列是 D 。 A.非周期序列 B.周期3=N C.周期8/3=N D. 周期24=N 2、一连续时间系统y(t)= x(sint)，该系统是 C 。 A.因果时不变 B.因果时变 C.非因果时不变 D. 非因果时变 3、一连续时间LTI 系统的单位冲激响应)2()(4-=-t u e t h t ，该系统是 A 。 A.因果稳定 B.因果不稳定 C.非因果稳定 D. 非因果不稳定 4、若周期信号x[n]是实信号和奇信号，则其傅立叶级数系数a k 是 D 。 A.实且偶 B.实且为奇 C.纯虚且偶 D. 纯虚且奇 5、一信号x(t)的傅立叶变换?? ?><=2||02||1)(ωωω， ， j X ，则x(t)为 B 。 A. t t 22sin B. t t π2sin C. t t 44sin D. t t π4sin 6、一周期信号∑∞ -∞ =-= n n t t x )5()(δ，其傅立叶变换)(ωj X 为 A 。 A. ∑∞ -∞ =-k k )52(5 2πωδπ B. ∑∞ -∞ =- k k )5 2(25πωδπ C. ∑∞ -∞ =-k k )10(10πωδπ D. ∑∞-∞ =- k k )10 (101 πωδπ 7、一实信号x[n]的傅立叶变换为)(ω j e X ，则x[n]奇部的傅立叶变换为
长沙理工大学拟题纸 课程编号 1 拟题教研室(或老师）签名 教研室主任签名 符号说明:)sgn(t 为符号函数，)(t δ为单位冲击信号,)(k δ为单位脉冲序列,)(t ε为单位阶跃信号,)(k ε为单位 阶跃序列。 一、填空(共3０分，每小题３分） １. 已知 )()4()(2 t t t f ε+=，求_______)("=t f 。)('4)(2)("t t t f δε+ 2. 已知}4,2,4,3{)(},1,2,2,1{)(=-=k h k f ,求______)()(=*k h k f 。}4,6,8,3,4,10,3{)()(-=*k h k f 3. 信号通过系统不失真的条件为系统函数_______)(=ωj H 。0 )(t j Ke j H ωω-= 4. 若)(t f 最高角频率为m ω,则对)4(t f 取样的最大间隔是______。 m T ωπωπ4max max == 5． 信号t t t f ππ30cos 220cos 4)(+=的平均功率为______。 10 1122222 =+++== ∑∞ -∞ =n n F P 6. 已知一系统的输入输出关系为)3()(t f t y =,试判断该系统是否为线性时不变系统 ______。故系统为线性时变系统。 ７. 已知信号的拉式变换为 )1)(1(1 )(2-+= s s s F ，求该信号的傅立叶变换)(ωj F =______。故傅立叶变 换)(ωj F 不存在。 ８． 已知一离散时间系统的系统函数 2121 )(---+= z z z H ，判断该系统是否稳定______。故系统不稳 定。 9. =+-+?∞ ∞-dt t t t )1()2(2δ______ 。3 1０. 已知一信号频谱可写为)(,)()(3ωωωω A e A j F j -=是一实偶函数，试问)(t f 有何种对称性______。关于ｔ=3的偶对称的实信号。 二、计算题（共50分,每小题1０分） 1． 已知连续时间系统的单位冲激响应)(t h 与激励信号)(t f 的波形如图A －1所示，试由时域求解该系 统的零状态响应)(t y ,画出)(t y 的波形。 图 A-1 1． 系统的零状态响应)()()(t h t f t y *=，其波形如图A －7所示。
信号与系统复习参考练习题一、单项选择题：
14、已知连续时间信号,)2(100) 2(50sin )(--=t t t f 则信号t t f 410cos ·)(所占有的频带宽度为（） A ．400rad ／s B 。200 rad ／s C 。100 rad ／s D 。50 rad ／s
f如下图（a）所示，其反转右移的信号f1(t) 是（） 15、已知信号)(t f如下图所示，其表达式是（） 16、已知信号)(1t A、ε（t）＋2ε(t－2)－ε(t－3) B、ε(t－1)＋ε(t－2)－2ε(t－3) C、ε(t)＋ε(t－2)－ε(t－3) D、ε(t－1)＋ε(t－2)－ε(t－3) 17、如图所示：f（t）为原始信号，f1(t)为变换信号，则f1(t)的表达式是（） A、f(－t+1) B、f(t+1) C、f(－2t+1) D、f(－t/2+1) 18、若系统的冲激响应为h(t),输入信号为f(t),系统的零状态响应是（）
19。信号)2(4sin 3)2(4cos 2)(++-=t t t f π π 与冲激函数)2(-t δ之积为（ ） A 、2 B 、2)2(-t δ C 、3)2(-t δ D 、5)2(-t δ ，则该系统是（）＞－系统的系统函数．已知2]Re[,6 51)(LTI 202s s s s s H +++= A 、因果不稳定系统 B 、非因果稳定系统 C 、因果稳定系统 D 、非因果不稳定系统 21、线性时不变系统的冲激响应曲线如图所示，该系统微分方程的特征根是（ ） A 、常数 B 、 实数 C 、复数 D 、实数+复数 22、线性时不变系统零状态响应曲线如图所示，则系统的输入应当是（ ） A 、阶跃信号 B 、正弦信号 C 、冲激信号 D 、斜升信号 23. 积分 ?∞ ∞-dt t t f )()(δ的结果为( ) A )0(f B )(t f C.)()(t t f δ D.)()0(t f δ 24. 卷积)()()(t t f t δδ**的结果为( ) A.)(t δ B.)2(t δ C. )(t f D.)2(t f