无人机三轴稳定云台的模糊PID控制

无人机控制系统中的模糊控制技术研究第一章绪论近年来，无人机作为新型的航空器，已经成为了军事、民用和商业领域中的重要应用。

与传统的飞机不同的是，无人机不需要人员操控，可以通过遥控器、计算机等设备进行操作。

无人机的控制系统不仅需要满足稳定性和安全性等基本要求，还需要考虑到其操作和控制的性能和效率。

模糊控制技术作为一种智能控制技术，在无人机控制系统中具有非常重要的应用价值。

本文主要研究无人机控制系统中的模糊控制技术，旨在分析其原理、特点及应用，为无人机控制系统的设计和实现提供理论指导和技术支持。

第二章模糊控制技术原理2.1 模糊控制技术概述模糊控制技术是一种基于模糊逻辑理论的智能控制技术，它能够模拟人类思维方式，通过对模糊量的量化处理来完成控制过程。

模糊控制技术通过对系统输入输出的关系进行模糊化，将输入输出之间的关系转化为一组人类语言的规则，再通过推理和模糊综合等方法来实现系统的控制。

2.2 模糊控制系统的基本结构模糊控制系统由模糊化、知识库、推理机以及去模糊化四部分组成。

其中模糊化是将实际输入转化为模糊量的过程，知识库包含了模糊控制的规则，推理机通过运用这些规则来得出控制量，去模糊化则是将模糊量转化为实际的控制量。

2.3 模糊集合及其运算在模糊控制系统中，模糊集合是一个重要概念。

模糊集合对于每一个元素都有一个隶属度，它表示了这个元素对于这个集合的模糊程度。

在进行模糊运算时，常用的有模糊交、模糊并、模糊补、模糊反等运算。

第三章无人机控制系统中模糊控制技术的应用无人机作为新型的飞行器，其控制系统需要精确的控制和运动性能。

而模糊控制技术恰好可以满足这一需求。

以下是无人机控制系统中典型的应用案例：3.1 无人机飞行控制系统中的模糊控制技术无人机飞行控制系统需要实现对飞行姿态的控制和稳定。

模糊控制技术可以优化控制器的设计和参数调节，使得飞行过程更加平稳和安全。

3.2 无人机障碍物避难系统中的模糊控制技术无人机障碍物避让系统需要根据传感器和图像信息，对周围环境进行感知和分析，实现障碍物避免和规避。

AUTONOMOUS QUADCOPTER STABILITYWITH PID CONTROLSri Wahyuni1, M. Izar Bahroni2, Faikul Umam3University of Trunojoyo MaduraMechatronics Department of Engineering FacultyEngineering FacultyBangkalan, East Java, Indonesia1**********************.id,2*******************,3*******************.idAbstract—Unmanned Aerial Vehicle (UAV) has many uses, including aerial photography, aerial mapping and monitoring activities. Quadcopter is a type of UAV that uses four rotors. The speed of each rotor has a considerable influence on the movement. The quadcopter movement can be done in this research, namely the process of arming, taking off, hover, and landing on decisions made by the system (autonomous). How the quadcopter achieves a balance in its movement to be stable and responsive requires a method. One method that is suitable for processing stability is the method (PID). The PID method has three main parameters, namely Proportional (Kp), Integral (Ki), Derivatives (Kd), with the determination of the constant through the trial and error process to obtain the optimal stability value as the purpose of this study. But the load of the quadcopter and the wind tightness around it is very influential to get the quadcopter movement to survive in stable conditions. Through a series of experimental processes carried out to produce the best constant values, at Kp = 1.3, Ki = 0.04, and Kd = 18, where the quadcopter is able to survive for 20 sec at the same relative point from the time of departure. The addition of GPS sensors in advanced research will be able to make this quadcopter move stable with the monitored position.Keywords—UAV; Quadcopter; PID; AutonomousI.I NTRODUCTIONQuadcopter / quadrotor is one type of UAV (Unmanned Aerial Vehicle) which has a manual or automatic control system. The development of this UAV has been used in various aspects. For example in agriculture. The Center for Agricultural Land Resources Research and Development began developing unmanned aircraft with the aim of analyzing the condition of plants / vegetation or agricultural land using VNIR, SWIR, thermal, radar or SAR bands [1]. Another example in the military field, the reconnaissance process in hard-to-reach areas is the background of the development of drones / UAVs by the National Aeronautics and Space Agency (LAPAN) [2].Aerial footage Photographty, Aerial footage Videography, Aerial mapping including activities that require drones / UAVs in the implementation. In addition to the skills of the user users, the stability and the surrounding natural conditions also have a considerable influence. A good drone is a drone / UAV that can maintain stability from outside interference, for example wind speed and frequency interference. Quadcopter is one type of UAV with four rotor drives. Each rotating rotor has a lift force and has the same distance to the center of gravity. This type of quadcopter is one type of drone that is quite easy to maneuver.The method used to be able to adjust the rotational speed of a quadcopter motor is the PID method. The method that uses proportional (Kp), integral (Ki), derivative (Kd) elements aims to accelerate the reaction of a system, eliminate offsets and produce large initial changes.Literatur ReviewResearchers of Gembong ES examined the height control on a quadcopter. In controlling the quadcopter, the PID method was used to obtain stability. Researchers are more likely to do testing in the form of simulations [3]. Likewise with Panca AK researchers, the method used is the Fuzzy PID in stabilizing the height of the quadcopter. This research only focuses on pitch and roll angle movements [4]. Guneshwor Singh menjelaskan tentang navigasi otomatis pada quadcopter.There are two types of propellers in the test. Propellers with a 10 inch size and a pitch of 4.5 can produce a greater total thrust than the use of an 8 inch size propeller with the same pitch size. This also affects flight time which tends to be longer when using a 10 inch propeller [5].Analysis of the stability of diagonal motion has been carried out by Salmaa. The analysis includes the combination of lateral motion and longitudinal motion in a quadcopter. Based on the analysis, the stability of the diagonal motion model is obtained that all diagonal motion models are unstable. Therefore PID control is needed by determining the gain of Kp, Ki, and Kd [6].Determination of the combination of PID constants is very influential on the control of stability of the quadcopter. The PID constant generated from the research conducted by Wili Kumara is Kp = 40, Ki = 40, Kd = 60 for roll motion while Kp = 40, Ki = 60, and Ki = 60 for pitch movement. However, the value of this constant has a fluctuation of 1 ° to (-5 °) slope [7].II.P ID M ETHODThe PID method is a control system consisting of three types of controllers, namely Proportional, Integral andInternational Conference on Science and Technology (ICST 2018)Derivative controllers. PID controller is one of the control methods that is often used in industrial control systems [8]. A. Proportional ControllerThe proportional controller has an output proportional to the magnitude of the error signal (the difference between the desired amount and the actual price). More simply it can be said that proportional controller output is a multiplication between proportional constants with input. Changes to the input signal will immediately cause the system to directly change its output by its constants.B. Integral ControllerThe integral controller functions to produce a system response that has a steady state error (zero steady state error). If a plant does not have an integrator element, the proportional controller will not be able to guarantee the system output with a zero steady state error. The use of an integral controller, the system response can be corrected, which has a zero steady state error.An integral controller has characteristics as well as an integral. The controller output is greatly affected by changes that are proportional to the error value. If the error signal does not change, the output will maintain the state as before the input changes.C. Derivative ControllerIntuitively, the oscillating system response is caused by several things. The dynamic process of a plant causes the response of a plant to not change immediately with a change in the control signal, but requires processing time. This time will make the control system experience delays to correct errors. This requires a controller that can predict errors from a system. Derivative controllers are generally used to speed up the initial response of a system, but do not minimize errors in the steady state. Derivative controller work is only effective in a narrow scope, namely in the transition period. Therefore the derivative controller is never used without any other controller of a system [8].III. BLOCK D IAGRAM S YSTEMThe block diagram in Fig. 1 is a system diagram consisting of several inputs which include the MPY-6050 type GY module in which there is an MPU-6050 sensor. Input in the form of slope data (gyrometer) and acceleration (accelerometer) becomes a reference for quadcopter balancing movements. The output of the system in the form of the speed value converted into a voltage value by ESC will drive a brushless motor that amounts to four. Brushless motor speed will be fully controlled using the PID method calculation. The ESC specification and brushless motor are adjusted to the load of the quadcopter.Input in the form of height sensor / wind pressure is a barometer and receiver from 2.4GHz radio control in the process on a different microcontroller. This section is as a medium for information on the height of the quadcopter when the quadcopter airs. Quadcopter communication with groundstation uses a bluetooth / wireless network module. The ground station receives altitude data from bluetooth to be able to control the 2.4Ghz transmitter radio control and provide information on the current mode status and height data on the personal computer.BMP-85MPU-6050Receiver (Rx)Arduino Nano Arduino Nano ESC 30A (4)Brushless MotorBluetoothPersonal ComputerTransmitterInputProcessOutputGround StationBluetoothFig. 1. Block Diagram SystemIV.F LOWCHARTA. General Flowchart Autonomous SystemIn Fig. 2 and Fig. 3 are general autonomous system flowcharts. Overall testing is done if the constant Kp, Ki, Kd is found to be optimal. This process is the whole process of moving the system on a quadcopter.Start Calibration Connecting ?NoYesCheck Bluetooth ConnectionSwitch ModeConnect PersonalComputerConnecting ?NoYesFig. 2. General flowchart of autonomous systems part 1Two bluetooth (master / slave) auto-pairing, both have ID identifiers, so bluetooth will be connected to bluetooth whose identification ID has been aligned. After the Bluetooth connection process is complete, the next step is to connect the personal computer to the system using a USB cable. Users can choose to use autonomous mode (automatic) or manual mode. When using autonomous mode, the system will run automatically according to the system design that has been made. The command to turn on the brushless motor is done by using a radio control transmitter. In the aeromodeling field, the process is often known as the Arming process. Then Take off is a quadcopter process to take off. Take off is done until it reaches a height of five meters. If the height has reached ten meters, the quadcopter will maintain height and stability for 30 seconds. Then the quadcopter will land / land slowly until it reaches the surface. But if the user uses manual mode. The manual program will be activated, all control systems are carried out manually by the user with a radio control device.Flowchart PID.Height = 10 m ?Arming/Motor ONMotor Off EndNoYes Take Off HoverLandingAutonomous Mode ON.?Start AutonomousProgram Start Manual ProgramYesNoCheck QuadcopterHeightControl by user (Transmitter)Fig. 3. General flowchart of autonomous systems part 2In Fig. 4 is a flowchart from the PID method. The PID method is used to maintain the stability of the quadcopter when moving. This method requires the main constants, Kp, Ki, Kd and Set point. The constant will be used to obtain an optimalsystem response. Determination of the value of Kp, Ki and Kd based on experiments by adjusting specifications on hardware. Starting with the reading of the gyro quadcopter value on the pitch, roll and yaw axis. The gyro data will be used as input for the calculation of the PID method. The output from the PID calculation will produce a PWM value for each brushless motor. The process will repeat the process until the error value is 0. The PID calculation is applied to the manual mode or autonomous mode.StartKp, Kd, Ki, Error, Last Error, SetpointRead MPU-6050 Gyro Value PID calculation Check ErrorError = 0EndYesNoPitch, Roll, Yaw PWM Motor BrushlessPWM motor1 = motor2 = motor3 = motor4Fig. 4. Flowchart PIDV.T ESTINGA.Bluetooth Distance TestingTesting is done in the open. This test aims to determine the range of the second Bluetooth communication. Because bluetooth communication is very important, the installation of the bluetooth module and the direction of the antenna on the bluetooth module must be considered very carefully. shown in Table 1.TABLE I.REACH TESTING DATAbluetooth communication, the average value of the range was 21.5 meters. If the Bluetooth range exceeds the range that has been entered in the test, the data communication will not be sent to the maximum or even thecommunication will be interrupted.B. Testing of Motor Lifting Power (Thrust)Lifting (thrust) testing is carried out on each brushless motor used on a quadcopter. Figure 5 is a graph of testing the lift using a 9540 (9 inch) propeller.Fig. 5. Motor Lift Power ChartC. Testing the PID MethodPID method is the main method used for quadcopter stabilization process. In testing this method, to get proportional (Kp), integral (Ki) and derivative (Kd) constants is done by trial and error. When the system is given a constant Kp, Ki, Kd is observed for quadcopter behavior. Testing the PID method is done in the test box. The test box is shown in Fig. 6.Testing to find a good system response is repeated. By observing the movement of the quadcopter in the stabilization process there will be some differences, including the response of the quadcopter stability to time. Here are some test results using Kp, Ki, Kd that have been done.Fig. 6. Stability TestFig. 7 is a test with a constant value of Kp = 1.3, Ki = 0.04 Kd = 5. The value of this constant is used continuously by the system. In addition to the response to steadystate fast enough, the response is not too visible to perform large oscillations. When testing, 200 sensor reading samples were taken with the assumption of data retrieval for four seconds.Fig. 7. Results of Testing the PID MethodVI. CONCLUSION AND FURTHER RESEARCHBased on the testing and analysis results that have been carried out in the research, the following conclusions are obtained:1. Quadcopter can perform optimal stability movementsusing the PID method. With a constant value of Kp = 1.3, Ki = 0.04 and Kd = 18.2. Delivery of altitude data sent via bluetooth communicationcan work well up to a distance of 21.5 meters.3. The initial position of the quadcopter during take off willaffect the shift of movement during flight.4.The type of propeller 1045 has a fairly large lift, but thetemperature of the motor when used is faster heat.5. The type of propeller 9450 tends to be more stable inmovement.6. Quadcopter tends to stay away from the initial coordinatepoint if the flight time is increased.From the results of the research conducted, for the next development, it is suggested that the following:1. Pay attention to the specifications of the quadcopter partused, especially in the specifications of the motor for lift strength and propeller type.munication for quadcopter and ground station shoulduse telemetry. So that the data transmission range can be carried out with a considerable distance and minimize frequency interference from outside.3.Adding GPS components, to determine the position /location when flying.4.Provides an electronic shock damper component that is inthe quadcopter to minimize the vibration generated by the four brushless rotors.5.Better mechanical design, to make it easier to implementthe control algorithm used.6.Pay attention to security factors during implementation.Addition of propeller protectors so that the propeller is safe when experiencing a system experiencing errors / crashes.7.When testing, you must first consider the state of the windspeed, this is to reduce the potential of the quadcopter carried by the wind (flyaway).A CKNOWLEDGMENTWe would like to show our gratitude to all our college and student in Mechatronics Department for sharing their wisdom during our research. This research was supported by Basic Mechatronic Laboratory.R EFERENCES[1]R. Shofiyanti, “Teknologi Pesawat Tanpa Awak untuk Pemetaan danPemantau an Tanaman dan Lahan Pertanian”, Informatika Pertanian, Vol.20, No.2, pp 58-64, 2011.[2]Aerostar TUAV : Drone Intai Andalan Skadron Udara 51 TNI AU.</aerostar-tuav-drone-intai-andalan-skadron udara-51-tni-au/> accessed on September 12th 2017.[3]G.E. Setyawan, E. Setiawan, W. Kurniawan, “Sistem KendaliKetinggian Quadcopter Menggunakan PID,” JT IIK, vol 2, no 2, pp 125-131, 2015.[4]P.A. Kusuma, A. Dharmawan, April, “Pengendalian KestabilanKetinggian pada Penerbangan Quadrotor dengan Metode PID Fuzzy,”IJE IS, Vol.7, No.1, pp 61-70, 2017.[5]I.O.G. Singh, “ Self-Navigating Quadcopter,” Int. J. Computer Sci. andInformation Technologies (IJCSIT), vol 6(3), pp 2761-2765, 2015.[6]Salmaa, “Analisa Kestabilan Gerak Diagonal pada Quadrotormenggunakan Kontrol PID,” Jurnal Ilmiah Matematika (MATH unesa), vol 3, no 6, 2017.[7]W.K. Juang, and L.L. Tung, “Pembuatan Mode l Quadcopter yang DapatMempertahankan Ketinggian Tertentu,” Jurnal Teknik Elektro, vol 9, no 2, 2016.[8]W.K. Juang, and L.L. Tung, “Pembuatan Mode l Quadcopter yang DapatMempertahankan Ketinggian Tertentu,” Jurnal Teknik Elektro, vol 9, no 2, 2016.。

分享无刷云台PID等参数说明以及调试教程参数说明：P，I，D - PID控制器设置，单独为每个轴P - 定义对干扰的响应。

越大，反应速度越快。

也可以解释为“增益”传感器信号之前，它会被传递给电机..如何调整：这个值从零慢慢增加，直至稳定的质量变得好。

太高的值可能导致系统自激（可见增加振荡）。

重要！如果从主框架的振动被传递到摄像机平台，它们可能导致简单的自我激励和不平衡。

在低频振荡的情况下，它们可以抑制通过增加D参数。

D的增加1 ..2个单位，如果将停止振荡 -你可能会进一步增加P。

D - 定义抑制反应。

它有助于防止低频振荡，但过高值可能会引入系统中的高频率的噪音的情况下，可能会增加振动。

尽量保持它尽可能接近零。

I - 定义反应的速度控制命令从接收器。

值给出非常低缓慢和平稳的操作员控制的反应，但也很缓慢的运动适当的地平线位置误差大的情况下。

高值让迅速移动相机和迅速返回到地平线，如有倾斜。

Power–定义对每个电机的输出功率。

设置从0到255，其中255 - 是最大的可用功率。

如何设置这个参数：过高的话会导致电机过热过低的话扭矩不够如果已经调过PID，还想再加power的话需要调小PIDInvert–定义了电机的旋转方向。

这是非常重要的，选择正确的设置。

那里自动检测。

要启动检测，设置P，I，D=0，设置功率上述建议（或一般值设置为100）。

级相机板水平（可能是不精确的，因为磁场矢量是随机的位置），然后按下自动按钮（位于附近设置反转）。

在校准过程中，相机将在小角度倾斜。

N.POLES -定义每个电机的极数（磁铁）。

此值大致在自动检测上面描述的校准程序。

如果检测到错误，手动：数一数磁铁电机（一般是14级）RC min.angleRC max.angle - 定义范围从最小的到最大角度的摄像头倾斜，从接收机控制。

逆控制，只需设置一个值大于第二个。

为例如，如果你想从地平线（0度）倾斜相机直降90度，min.angle=0，max.angle= 90（或90 ..逆为0）。

模糊PID算法的优化及其在无人机中的应用因结构简单、操作方便、机动性强、能垂直起降等特点,四旋翼无人机已成为工业界和学术界的热门研究课题。

但四旋翼无人机对环境变化极为敏感,使得对飞行姿态的精确控制难度较大,这对无人机飞控系统的设计者来说是一个严峻挑战,也使得飞行控制算法成为对四旋翼无人机控制的关键技术之一。

到目前为止,四旋翼无人机的核心控制方法仍然是PID控制,因为该飞行控制方法不需要对系统进行精确建模,只需根据系统误差,通过模糊规则调整控制器的参数,实现飞行控制。

基于国内外现有的研究成果,本论文对无人机PID控制器的设计进行了改进和测试,主要研究内容包括如下三个方面:1)对无人机姿态角修正方法的研究。

采用卡尔曼滤波算法降低传感器输出噪声对姿态角的干扰,从而提高姿态计算的准确性,通过MATLAB仿真以及实际试飞验证了卡尔曼滤波算法的有效性。

实验表明无人机姿态角的实际值和卡尔曼滤波算法估计值的仿真曲线高度相似,证实了该姿态角修正方法可显著提高无人机飞行控制系统输入值的可靠性;2)对无人机的姿态控制算法的研究。

设计方法结合经典PID和模糊控制的优点,通过选择合适的模糊规则建立参数自整定的模糊PID控制器,实现了PID参数的自行校正,增强了无人机对周围环境的适应性。

理论分析表明,模糊PID控制器控制精度和抗干扰能力均有所提升,但与传统PID控制器相比,稳态误差更大。

同时对无人机飞控系统动态特性及噪声来源进行了分析,对PID控制器进行特定应用环境下的优化,并利用MATLAB搭建出飞控系统的软件仿真平台,对优化后的控制器进行了效果验证;3)对无人机实验平台搭建的研究。

为验证优化后PID控制器对无人机实际控制的有效性,在论文第四章进行了无人机软硬件平台的搭建及实际试飞测试。

该章节描述了实验平台的硬件模块功能和驱动电路,软件设计部分给出了无人机控制的算法流程图和传感器采集数据及数据处理的算法流程图。

在实际试飞过程中,无人机的姿态信息通过无线回传至遥控器,遥控器通过串口可将姿态信息转发至PC机。

无人机三轴稳定云台的模糊PID控制郭炳坤，陈水利，吴云东，李渭（集美大学理学院，福建厦门 361021）［摘要］以无人机三轴稳定云台的内框作为研究对象，将自适应卡尔曼滤波算法与模糊PID控制算法相结合，提出了一种基于自适应卡尔曼滤波的模糊PID控制算法.经过Matlab仿真实验表明，相对于经典PID控制算法和模糊PID控制算法而言，该算法在无人机三轴稳定云台的控制上，不仅响应速度快、精度高，而且对控制干扰噪声和测量噪声也起到了较好的抑制作用.[关键词]无人机；三轴云台；自适应控制；模糊PID；模糊规则；自适应卡尔曼滤波Fuzzy-PID Control for UAV Three-axis Stable Pan-tiltGUO Bing-kun,CHEN Shui-li,WU Yun-dong,LI Wei(School of Science,Jimei University,Xiamen 361021,China)Abstract:With its research focus on the inner frame of UAV three-axis stable pan-tilt,combining the adaptive Kalman filter algorithm and the fuzzy PID control algorithm,the fuzzy PID control algorithm based on adaptive Kalman filter was presented.Through the simulation using Matlab,the results show that this algorithm have an more advantage than the classical PID control algorithm and fuzzy PID control algorithm in UAV three-axis stable pan-tilt.It has fast response and high precision.The system also has a good inhibitory effect on control interference noise and measure noise.Key words:UAV；three-axis pan-tilt；self-adaptive control；fuzzy-PID；fuzzy rule；Adaptive Kalman filter。

飞行器控制中的模糊PID控制策略研究飞行器控制是一项非常重要的技术，它对于安全、舒适的飞行和有效的任务完成起着关键作用。

控制飞行器的关键是要保持其飞行姿态的稳定。

而PID控制器是其中一种常用的控制策略，它可以通过反馈控制来将系统保持在目标状态，但是在某些情况下，PID控制器更难以处理，此时就需要使用模糊PID控制策略来解决问题。

一、PID控制器的优缺点PID控制器被广泛应用于飞行器控制中，其优点在于简单易用，拥有很好的干扰鲁棒性和稳定性。

在PID控制器中，输出产生的偏差经过比例、积分、微分三个环节得出控制量，使得飞行器能够保持目标状态。

虽然PID控制器具有这些优点，但是其在某些情况下也有缺点。

例如，PID控制器的响应速度较慢，容易因参数调节不当而产生过调和欠调等问题。

因此，它并非在所有情况下都能够产生理想效果。

二、模糊PID控制器的原理模糊PID控制器是一种PID控制器的改进。

它可以通过模糊逻辑控制的方法来对PID控制器进行优化。

模糊PID控制器的原理是把PID控制器的三个环节的控制误差和变化率转化为模糊度量，然后使用模糊逻辑控制器进行处理，最后再转化为输出量。

模糊PID控制器可以分为两个部分：一是前馈回路，用于测量控制器输入量的变化率；二是反馈回路，用于根据控制器输入量和目标输出量的误差进行调节。

三、模糊PID控制器的优缺点相比于PID控制器，模糊PID控制器具有更好的鲁棒性、稳定性和响应速度。

这是因为模糊PID控制器能够通过控制静态误差和变化率来调整输出量，从而实现更精确的控制。

此外，模糊PID控制器还能够自适应环境和参数变化，有效应对不确定性问题。

然而，模糊PID控制器相较于PID控制器也存在一些缺点。

首先，模糊PID控制器需要复杂的算法，并且调试参数是非常困难的。

其次，模糊PID控制器需要进行系统建模，需要消耗更多的计算资源。

四、应用举例模糊PID控制器的在飞行器控制中的应用，可以举出一些例子。