Neural Network PID Research Articles

Research and Design of Precision Fertilizer Application Control System Based on PSO-BP-PID Algorithm

China has had the highest fertilizer use rate in the world for years, but today a large number of farmlands still use traditional manual fertilizer application methods, which rely entirely on personal experience and not only cause the waste of fertilizer and water resources but also make the local ecological environment polluted. This paper researches and designs a BP neural network PID controller based on PSO optimization to address the above problems. The PSO algorithm is used to optimize the initial weights of the BP neural network, and then optimize the control parameters of the PID to achieve accurate control of the liquid fertilizer flow. A precision fertilizer control system based on the STM32 microcontroller was also developed, and the performance of this controller was verified in tests. The results showed that compared with the conventional PID controller and BP neural network-based PID controller, this controller had good control accuracy and robustness, the average maximum overshoot was 6.35%, and the average regulation time was 41.17 s; when the fertilizer application flow rate was 0.6 m3/h, the shortest adjustment time is 30.85 s, which achieves the effect of precise fertilizer application.

Adaptive PI and RBFNN PID Current Decoupling Controller for Permanent Magnet Synchronous Motor Drives: Hardware-Validated Results

This study presents an adaptive proportional-integral (PI) and radial basis function neural network proportional-integral-derivative (PID) current control solution for permanent magnet synchronous motor (PMSM) drives. The proposed controller includes four controls: a decoupling term, a PI term, a supervision term, and a radial basis neural network-PID (RBFNN-PID) term. The first control term makes up the nonlinear factors, the second automatically adjusts the control gains, the third guarantees the system stability, and the fourth optimizes the PID parameters to achieve optimal system performance. Unlike off-line tuned PID controllers, the adaptive controller includes an adaptive tuning method for the on-line adjustment of control gain based on the gradient descent strategy. Therefore, it can be adjusted and handle the uncertainty of any system parameters. The program is not only simple and easy to be implemented but also ensures the accuracy and rapidity of the tracking speed. The control system has proven to be asymptotically stable. To verify the theory and application of the algorithm, a comparative experiment between the adaptive PI + RBFNN-PID controller and traditional PI + PID controller was conducted, demonstrating good robustness, which can effectively improve the PMSM. The results confirm that the proposed design achieves excellent control stability (i.e., quicker transient responding and smaller steady state error) in the presence of parametric uncertainty compared to conventional PID methods.

3-DOF Position and Orientation Control of an Air Flotation Platform for Spacecraft Ground Microgravity Simulation by Using Double Closed-Loop Cascade PIDnn

Space assistant robots can help astronauts or their assistants perform certain tasks. A ground microgravity simulation environment is built for the space assistant robot AAR-2. The hardware requirements of the ground simulation by the 3-DOF microgravity air flotation platform. An algorithm is designed for this simulation system. By using momentum and RMSprop methods to improve the PID neural network, the challenging problem of strong coupling between system nonlinearity and variables is solved. Firstly, the paper introduces the hardware system and deduces the dynamic model of the system. Then, the algorithm is calculated and simulated. Through simulation, the effectiveness and feasibility of the algorithm are compared and proved. Finally, the control system is simulated by MATLAB/Simulink and compared with other advanced algorithms. The simulation results show that the designed neural network controller can quickly and accurately control the 3-DOF of freedom motion of AAR-2.

Study on Early Warning of Karst Collapse Based on the BP Neural Network

In order to comprehensively grasp the dynamics of karst collapse, promote the comprehensive prevention and control level of karst collapse, and prevent secondary disasters caused by lava collapse, this study presents a method of karst collapse early warning based on the BP neural network. This method does not need to set the sliding surface in the finite element calculation model. The stress of the sliding surface is fitted according to the spatial stress relationship of the deep karst layer through the improved BP neural network PID control algorithm and BP neural network algorithm, which avoids the modeling and mesh generation of the complex sliding block and has good accuracy and ease of use. According to the basic theory of the BP neural network, the calculation formulas of multilayer feedforward and error back propagation processes are derived, and the two-dimensional and three-dimensional finite element models of gravity dams without and with sliding blocks are established, respectively. Finally, according to the common formulas of viscoelastic artificial boundary and equivalent load, the two-dimensional and three-dimensional input programs of the karst fluid state are compiled, and a neural network early warning model is obtained. The experimental results show that the process karst state simulated by the algorithm is very close to the actual situation, and the minimum value of antisliding coefficient and its occurrence time can be accurately predicted, with an error range of less than 3%. Conclusion. BP neural network prediction can effectively predict karst collapse, with higher prediction accuracy, and can effectively simulate the actual collapse risk.

Application of PSO-BPNN-PID Controller in Nutrient Solution EC Precise Control System: Applied Research.

In this paper, we present a nutrient solution control system, designing a nutrient solution electrical conductivity (EC) sensing system composed of multiple long-range radio (LoRa) slave nodes, narrow-band Internet of Things (NB-IoT) master nodes, and a host computer, building a nutrient solution EC control model and using the particle swarm optimization (PSO) algorithm to optimize the initial weights of a back-propagation neural network (BPNN). In addition, the optimized best weights are put into the BPNN to adjust the proportional–integral–derivative (PID) control parameters Kp, Ki, and Kd so that the system performance index can be optimized. Under the same initial conditions, we input EC = 2 mS/cm and use the particle swarm optimization BP neural network PID (PSO-BPNN-PID) to control the EC target value of the nutrient solution. The optimized scale factors were Kp = 81, Ki = 0.095, and Kd = 0.044; the steady state time was about 43 s, the overshoot was about 0.14%, and the EC value was stable at 1.9997 mS/cm–2.0027 mS/cm. Compared with the BP neural network PID (BPNN-PID) and the traditional PID control approach, the results show that PSO-BPNN-PID had a faster response speed and higher accuracy. Furthermore, we input 1 mS/cm, 1.5 mS/cm, 2 mS/cm, and 2.5 mS/cm, respectively, and simulated and verified the PSO-BPNN-PID system model. The results showed that the fluctuation range of EC was 0.003 mS/cm~0.119 mS/cm, the steady-state time was 40 s~60 s, and the overshoot was 0.3%~0.14%, which can meet the requirements of the rapid and accurate integration of water and fertilizer in agricultural production.

Fuzzy Neural Network PID Strategy Based on PSO Optimization for pH Control of Water and Fertilizer Integration

In the process of crop cultivation, the application of a fertilizer solution with appropriate pH value is more conducive to the absorption of nutrients by crops. If the pH of the irrigation water and fertilizer solution is too high, it will not only be detrimental to the absorption of nutrients by the crop, but will also damage the structure of the soil. Therefore, the precise regulation of pH in water and fertilizer solutions is very important for agricultural production and saving water and fertilizer. Firstly, the article investigates the hybrid control of fertilizer and water conditioning systems, then builds a fuzzy preprocessing controller and a neural network proportional–integral–differential controller, and optimizes the neural network parameters by means of an improved particle swarm algorithm. The effectiveness of the controller was verified by comparison with the common proportional–integral–differential control and fuzzy algorithm control for fertilizer control and fuzzy preprocessing neural network control. Simulation experiments for this study were designed through the MATLAB/Simulink simulation platform, and the simulation results show that the algorithm has good tracking and regulation capabilities in the system. Finally, the four control algorithms are experimentally validated under different pH regulations using designed field experiments. The results show that, compared with other control algorithms, the control algorithm in this paper has a smaller overshoot and good stability with a shorter rise time, which can achieve the purpose of better regulating the fertilizer application system.

Face recognition and real-time tracking system based on convolutional neural network and parallel-cascade PID controller

The purpose of this research is to develop a high-efficiency, low-cost, and easy-to-use tracking system for vehicles, and it is expected that the system can be extended to areas such as service robots, autonomous driving, and manufacturing. In this paper, we introduced an object detection algorithm based on convolutional neural networks to realize face recognition, which has better efficiency and robustness than traditional machine learning methods. With the concept of edge computing, we deployed the model on the local embedded system to improve the information transmission and security issues of cloud computing. In order to realize the tracking system, this paper builds a mecanum-wheel vehicle with omnidirectional mobility, and proposes a parallel-cascade PID controller architecture based on the mecanum-wheel vehicle. The fixed distance linear tracking control can be realized through the dual-loop feedback control of distance and yaw angle; moreover, the vehicle slipping which is caused by difference rotation speed can be improved. Finally, through algorithm optimization, controller parameter adjustment, and system integration, an omnidirectional mobile vehicle with recognition and tracking functions is realized. The experiment results indicate that the system is stable and robust during actual operation.

Performance Comparison of PID and NARX Neural Network for Attitude Control of a Quadcopter UAV

ABSTRACT: In this study, two different types of controllers have been designed and tested for altitude and motion control of an autonomous quadrotor to compare the control performance under the influence of parametric uncertainty and disturbances. The first controller is a proportional-integral-derivative (PID) controller which is a conventional linear controller. The closed-loop PID algorithms calculate the results of the system by using the error values that consist of the difference between the sensor values measured by the closed-loop feedback method and the reference inputs. The second method that has been used is artificial neural network (ANN) algorithms, which provide both advantages and convenience in defining and controlling linear systems and non-linear systems with the closed-loop feedback method used in PID. The most important feature of the ANN algorithms is their high performance as a result of training with different input values. Therefore, the ANN control system has been trained with the input data used with Gaussian noise and the desired target data. A dynamic time series non-linear autoregressive with Exogenous input (NARX) neural network has been chosen as an ANN controller because of the time-delayed backpropagation learning performance. In this study, PID, and NARX NN control algorithms to control the maneuvers and altitude of the quadcopter and the mathematical model have been designed on Matlab Simulink. Motion control performances of the PID and NARX controllers are tested on the model. The design was tested on a real-time simulation environment with a one-millisecond fixed-step size. This paper proposes an alternative approach to control attitude and altitude on a quadcopter with the NARX NN algorithm.

Design of Distributed Drive Control Strategy for Wheel Side Rear Drive Electric Bus Based on Neural Network Algorithm.

The electric bus driven by the wheel-side motor has a natural advantage for the driving stability control of the vehicle because the torque of each driving wheel is independently controllable. Based on the advantages of the artificial neural network control algorithm, this paper designs a direct yaw torque control strategy based on neural network PID control, and combines the steering characteristics of the vehicle to distribute the driving torque. The co-simulation results of Matlab/Simulink and TruckSim show that the designed control strategy can effectively reduce the vehicle’s center of mass slip angle and lateral acceleration under medium and high speed conditions, and ensure the stable driving of the bus.

A Summary of PID Control Algorithms Based on AI-Enabled Embedded Systems

Proportional-integral-derivative (PID) controllers are extensively used in engineering practices for their simple structures, robustness to model errors, and easy operations. At present, there is a great variety of PID controllers. Companies have developed intelligent regulators with functions for automatically tuning PID parameters. For present PID controllers, strategies such as intelligence, self-adaptation, and self-correction are extended to transmission PID. PID controllers and corresponding improved ones are utilized in 90% of industrial control processes. In this paper, PID control algorithms are summarized. This paper focuses on advanced control strategies such as PID control, predictive PID control, adaptive PID control, fuzzy PID control, neural network PID control, expert intelligent PID control, PID control based on genetic algorithms, and PID control based on ant colony algorithms. Besides, these kinds of algorithms are compared, and prospects of PID algorithms are forecast at the end of this paper.

Trajectory tracking of differential drive mobile robots using fractional-order proportional-integral-derivative controller design tuned by an enhanced fruit fly optimization

This work proposes a new kind of trajectory tracking controller for the differential drive mobile robot (DDMR), namely, the nonlinear neural network fractional-order proportional integral derivative (NNFOPID) controller. The suggested controller’s coefficients comprise integral, proportional, and derivative gains as well as derivative and integral powers. The adjustment of these coefficients turns the design of the proposed NNFOPID control further problematic than the conventional proportional-integral-derivative control. To handle this issue, an Enhanced Fruit Fly Swarm Optimization algorithm has been developed and proposed in this work to tune the NNFOPID’s parameters. The enhancement achieved on the standard fruit fly optimization technique lies in the increased uncertainty in the values of the initialized coefficients to convey a broader search space. subsequently, the search range is varied throughout the updating stage by beginning with a big radius and declines gradually during the course of the searching stage. The proposed NNFOPID controller has been validated its ability to track specific three types of continuous trajectories (circle, line, and lemniscate) while minimizing the mean square error and the control energy. Demonstrations have been run under MATLAB environment and revealed the practicality of the designed NNFOPID motion controller, where its performance has been compared with that of a nonlinear Neural Network Proportional Integral Derivative controller on the tracking of one of the aforementioned trajectories of the DDMR.

Simulation and Experimental Study of Nonlinear Characteristics for Multi-mode Driving of Intelligent Vehicles

To study the nonlinear characteristics simulation and experiment of multimode driving of intelligent vehicle, an improved neural network PID controller is proposed to realize the safe and active braking requirements of cruise vehicles, according to the study of vehicle dynamics of different active braking forms and the joint simulation of road condition, driving condition, and braking mechanical structure. On the basis of the current research, the four-layer neural network combined with PID control is proposed in this paper, which can better control the braking compared with the traditional PID controller and BP neural network, safety, efficiency and stability have been improved in the following cars. According to the animation simulation demonstration of CAR SIM software, it can be determined that the improved neural network PID controller in the mechanical system can realize the safety and active braking requirements of cruise vehicle, and it is more accurate and efficient than the traditional BP neural network PID controller. Relevant parameters of the existing millimeter-wave radars are set by the radar sensors, and the long-distance horizontal detection range is plus or minus 10 degrees. The radar is installed on the longitudinal axis of the front bumper and has a height of 442[Formula: see text]mm. According to Car SIM’s preset radar sensor module parameters, the sensors can independently collect and transmit relevant data, it is also possible to cooperate with other sensors for the same variable parameter to increase the accuracy and avoid abnormal errors caused by the failure of a certain sensor. Compared with the traditional PID controller and BP neural network, the combination of four-layer neural network and PID control can carry out better braking control and improve the safety, efficiency, and stability of the following cars.

A Design of FPGA-Based Neural Network PID Controller for Motion Control System.

In the actual industrial production process, the method of adaptively tuning proportional–integral–derivative (PID) parameters online by neural network can adapt to different characteristics of different controlled objects better than the controller with PID. However, the commonly used microcontroller unit (MCU) cannot meet the application scenarios of real time and high reliability. Therefore, in this paper, a closed-loop motion control system based on BP neural network (BPNN) PID controller by using a Xilinx field programmable gate array (FPGA) solution is proposed. In the design of the controller, it is divided into several sub-modules according to the modular design idea. The forward propagation module is used to complete the forward propagation operation from the input layer to the output layer. The PID module implements the mapping of PID arithmetic to register transfer level (RTL) and is responsible for completing the output of control amount. The main state machine module generates enable signals that control the sequential execution of each sub-module. The error backpropagation and weight update module completes the update of the weights of each layer of the network. The peripheral modules of the control system are divided into two main parts. The speed measurement module completes the acquisition of the output pulse signal of the encoder and the measurement of the motor speed. The pulse width modulation (PWM) signal generation module generates PWM waves with different duty cycles to control the rotation speed of the motor. A co-simulation of Modelsim and Simulink is used to simulate and verify the system, and a test analysis is also performed on the development platform. The results show that the proposed system can realize the self-tuning of PID control parameters, and also has the characteristics of reliable performance, high real-time performance, and strong anti-interference. Compared with MCU, the convergence speed is far more than three orders of magnitude, which proves its superiority.

Neural network PID control for combustion instability

Considerable research studies have been reported on developing effective active control means to mitigate combustion instability. In this work, a neural network PID (NN-PID) controller is proposed and demonstrated to suppress the oscillating pressure in a cylindrical Rijke tube, for which the combustion instability is modelled by a 1D acoustic network model together with linear and nonlinear flame models. The flame model is based on the classical n − τ model filtered by a first-order low pass filter. The nonlinearity of the flame model is realised using a gain function saturated with velocity perturbation. A fuel valve and a loudspeaker are adopted as two independent control actuators. Results show that system oscillation begins to attenuate when the NN-PID control starts. The attenuation rate using a fuel valve as the actuator is faster than that using a loudspeaker. For nonlinear combustion oscillations, the NN-PID controller could inhibit the reference pressure oscillation in the linear growth and saturation processes. A maximum overshoot of 2.3% could occur when the loudspeaker is used as the actuator to suppress system oscillation. The NN-PID controller is not affected by noise with a fuel valve or the loudspeaker and can effectively suppress and eliminate the system pressure oscillation in different oscillating stages with different flame models.

Deep PID Neural Network Controller for Precise Temperature Control in Plastic Injection-moulding Heating System

In this paper, deep PID neural network (PIDNN) controller is used in a nonlinear thermal regulatory control problem. We have a hot runner system and a nozzle with round tip as controlled plants. The control goal is to regulate their temperature responses for tracking a constant set point stably and precisely without exactly knowing mathematic models of plants during entire process by using deep PIDNN controller. The parameters (or weights) of controller are updated on-line based on gradient descent rule with Adam optimizer. Comparing to the results with PID control, deep PIDNN controller reduces overshoot, saves much power and enhance the control performance that temperatures almost fluctuate within ±0.2°C tolerance of set point in steady state.

Research on DC Electronic Load System Based on BP Neural Network PID Control

DC electronic load is a kind of instrument to replace the traditional dissipative resistance, which is widely used in the discharge performance test of DC power supply. However, it is difficult to achieve better control effect through analog circuit control in some test environments with high speed requirements for discharge test, and it is easy to produce overshoot. In digital circuits, however, in the actual control of DC electronic load, there are many uncertainties in the control system, such as interference and delay, which makes it difficult for the system to achieve the best control effect. This paper takes constant current mode as an example, in order to improve the effect of PID control, BP (Back Propagation) neural network is used to dynamically adjust the PID parameters online. After simulation analysis, under the ideal control environment, the adjustment time of BP-PID is only 0.0002s, which is 0.0005s higher than the traditional PID of 0.0007s. It can also reach a stable value within 0.0003s in the interference environment, with almost no overshoot. Therefore, compared with traditional PID control or analog circuit control, BP neural network PID control has better control effect on DC electronic load system.

Adaptive Fuzzy Neural Network PID Algorithm for BLDCM Speed Control System

Because of its simple structure, high efficiency, low noise, and high reliability, the brushless direct current motor (BLDCM) has an irreplaceable role compared with other types of motors in many aspects. The traditional proportional integral derivative (PID) control algorithm has been widely used in practical engineering because of its simple structure and convenient adjustment, but it has many shortcomings in control accuracy and other aspects. Therefore, in this paper, a fuzzy single neuron neural network (FSNNN) PID algorithm based on an automatic speed regulator (ASR) is designed and applied to a BLDCM control system. This paper introduces a BLDCM mathematical model and its control system and designs an FSNNN PID algorithm that takes speed deviation e at different sampling times as inputs of a neural network to adjust the PID parameters, and then it uses a fuzzy system to adjust gain K of the neural network. In addition, the frequency domain stability of a double closed loop PID control system is analyzed, and the control effect of traditional PID, fuzzy PID, and FSNNN PID algorithms are compared by setting different reference speeds, as well as the change rules of three-phase current, back electromotive force (EMF), electromagnetic torque, and rotor angle position. Finally, results show that a motor controlled by the FSNNN PID algorithm has certain superiority compared with traditional PID and fuzzy PID algorithms and also has better control effects.

GRNN inverse system based decoupling control strategy for active front steering and hydro-pneumatic suspension systems of emergency rescue vehicle

Aiming at eliminating the mutual interference and coupling between the active front steering (AFS) and hydro-pneumatic suspension (HPS) subsystems of the emergency rescue vehicle, this paper presents a neural network inverse system based decoupling control strategy, which consists of the neural network inverse system and PID feedback control strategy. At first, a three-degrees of freedom dynamics model of the vehicle is constructed, and the reversibility of the vehicle chassis system is analyzed by the Interactor algorithm. Then, the generalized regression neural network (GRNN) model is used to identify the vehicle chassis inverse system, which decouples the multivariable vehicle chassis system into two independent single input and single output systems. Next, this paper proposes a closed-loop compound controller for the single input and single output systems, which is designed based on the PID control strategy and GRNN inverse system to realize the feedback control of the vehicle chassis system, so that the vehicle can track the desired yaw rate and roll angle well. Simulations and experiments are conducted on the self-developed fire rescue prototype vehicle, and the results prove that the proposed decoupling control strategy can not only eliminate coupling between AFS and HPS subsystems of the vehicle chassis, but also make the vehicle have better handling stability than controlled by the integrated control.

Fuzzy Neural Network PID Control of Valve-controlled Hydraulic Motor for Camellia Fruit Picking Machine

Fuzzy Neural Network PID Control of Valve-controlled Hydraulic Motor for Camellia Fruit Picking Machine

Optimal torque distribution of distributed driving AGV under the condition of centroid change

The torque distribution is researched under the condition of the centroid position of distributed drive automatic guided vehicle (AGV) with load platform and is uncertain due to the unknown movable load. The whole vehicle model under centroid variation, the efficiency model of the hub motor and the torque distribution control strategy based on a PID neural network are established. A hierarchical controller is designed to accurately ensure the economy and stability of the vehicle. Simulations of the proposed control strategy are conducted, the results show that the total power and lateral deviation distance of the driving wheels are reduced by 17.63% and 61.54% under low load conditions and 15.54% and 61.39% under high load conditions, respectively, compared with those of the driving wheels under the average torque distribution, and the goal of close slip rates of the driving wheels is achieved. A system prototype is developed and tested, and the experimental results agree with the simulation within error permissibility. The margin of error is less than 5.8%, the results demonstrate that the proposed control strategy is effective. This research can provide a theoretical and experimental basis for the torque optimization distribution of distributed drive AGVs under centroid variation conditions.