AGV path tracking and optimization method based on DSP-based neural network algorithm, computer storage medium and device

Abstract

The application relates to an AGV path tracking and optimization method based on a neural network algorithm of a DSP, a computer storage medium and equipment, and aims to solve the problems that the existing AGV path planning algorithm is prone to be interfered by the environment, has a slow convergence speed and is prone to fall into a local optimal state. The application comprises the following steps: 1) establishing a motion state mathematical model of AGV; 2) establishing a servo control system mathematical model of AGV; 3) constructing a neural network and learning to obtain a neural network training result; and 4) optimizing the motion state mathematical model of AGV and the servo control system mathematical model of AGV according to the neural network training result, and performing simulation test.

Description

Technical Field

[0001] This invention relates to a path optimization method, specifically to an AGV path tracking and optimization method based on a DSP-based neural network algorithm, a computer storage medium, and a device. Background Technology

[0002] Automated Guided Vehicles (AGVs) are battery-powered, driverless automated transport vehicles equipped with non-contact control and guidance systems. With the gradual development of factory automation and intelligent manufacturing, AGVs are widely used due to their high efficiency and flexibility, and can, to some extent, replace manual labor, reduce labor costs, and improve production efficiency. However, in limited usage scenarios or environments, and with increased numbers of AGVs or tasks to be performed, path planning for each AGV is crucial to ensure it reaches its target point without encountering static obstacles or dynamic conflicts. This involves considering the temporal position of each AGV and resolving route conflicts between multiple AGVs. If commands change, such as temporarily added tasks or AGVs deviating from their expected paths, these centralized solutions may need to replan all AGV tasks and routes. In such cases, the AGVs cannot quickly switch to the new command execution state, causing time delays and path conflicts. Therefore, AGV path planning is currently a hot research topic.

[0003] Path planning refers to the optimal path planning problem for robots, which involves finding the optimal path from the initial state to the target state while avoiding obstacles, based on one or more optimization criteria (such as minimizing work cost, shortest travel path, and shortest travel time). The applications of robot path planning are extremely diverse, ranging from the movement of NPCs and control characters in games, navigation problems in Baidu Maps, to the cleaning routes of household robot vacuums and the flight paths of drones, to the autonomous driving paths of self-driving cars that companies are vying to develop. For the path planning problem of AGVs (Automated Guided Vehicles), the mainstream algorithms currently include the A* algorithm and standard intelligent optimization algorithms. The A* algorithm, as an extension of Dijkstra's algorithm, is widely used for pathfinding and graph traversal due to its efficiency. However, when multiple minima exist, the A* algorithm cannot guarantee the optimal search path. Furthermore, both algorithms are susceptible to environmental interference, exhibiting slow convergence speeds and a tendency to get trapped in local optima. Summary of the Invention

[0004] The purpose of this invention is to solve the problems of existing AGV path planning algorithms being susceptible to environmental interference, having slow convergence speed, and being prone to getting trapped in local optima, and to provide an AGV path tracking and optimization method, computer storage medium, and device based on a DSP-based neural network algorithm.

[0005] To achieve the above objectives, the technical solution provided by this invention is:

[0006] A method for AGV path tracking and optimization based on a DSP-based neural network algorithm, characterized by the following steps:

[0007] Step 1: Establish a mathematical model of the AGV's motion state, obtain the motion state equations of the AGV in the two-dimensional plane at time t and at time t+Δt, and calculate the observation equation Z of the AGV relative to the target point based on the motion state equations of the AGV in the two-dimensional plane at time t and at time t+Δt.

[0008] Step 2: Establish a mathematical model of the servo control system of the AGV and obtain the total transfer function G(s) of the servo control system;

[0009] Step 3: Based on the observation equation Z of the AGV relative to the target point and the total transfer function G(s) of the servo control system, construct a neural network and learn it until the error reaches the preset requirement or the number of learning iterations reaches the preset requirement, and obtain the neural network training results.

[0010] Step 4: In the computer software, call the neural network training results obtained in Step 3 to optimize the observation equation Z of the AGV relative to the target point and the total transfer function G(s) of the servo control system, to obtain the optimized observation equation Z1 of the AGV relative to the target point and the optimized total transfer function G1(s) of the servo control system. Then, call the neural network training results obtained in Step 3, and perform simulation tests based on Z1 and G1(s) to obtain the test results. If the test results do not meet the preset requirements, update Z and G(s) using Z1 and G1(s), return to Step 3, and rebuild the neural network for training. If the test results meet the preset requirements, the path planning of the AGV is completed.

[0011] Furthermore, in step 1, the motion state equation of the AGV in the two-dimensional plane at time t is:

[0012]

[0013] Where t is time, x(t) and y(t) are the x-direction and y-direction positions of the AGV at time t, respectively, u(t) is the angle between the AGV's motion direction at time t and the x-axis, v(t) is the velocity of the AGV at time t, ω(t) is the angular velocity of the AGV at time t, and a(t) is the acceleration of the AGV at time t. d is the differential symbol;

[0014] At time t+Δt, the motion state equation of the AGV in the two-dimensional plane is:

[0015]

[0016] Among them, V max and V min These represent the maximum and minimum speeds of the AGV between time t and t+Δt, respectively.

[0017] At time t, the observation equation Z of the AGV relative to the target point is:

[0018]

[0019] Where D represents the distance between the current position of the AGV and the target point. x is the azimuth angle of the AGV's current position relative to the target point. d and y d Let x and y represent the position coordinates of the target point, respectively. t and y t Let x and y represent the position coordinates of the AGV at time t, respectively.

[0020] Further, in step 2, the mathematical model of the servo control system includes a servo driver, a servo motor, a speed reducer, and a controlled object connected in sequence, as well as a servo controller, a current feedback module, a speed feedback module, an attitude feedback module, a first arithmetic unit, and a second arithmetic unit; the first input terminal of the first arithmetic unit is connected to the input terminal of the AGV servo control system, and the output terminal is connected to the input terminal of the servo controller; the output terminal of the servo controller is connected to the first input terminal of the second arithmetic unit; the input terminal of the current feedback module is connected between the servo motor and the speed reducer, and the output terminal is connected to the second input terminal of the second arithmetic unit; the output terminal of the second arithmetic unit is connected to the input terminal of the servo driver; the input terminal of the speed feedback module is located between the speed reducer and the controlled object, and the output terminal is connected to the second input terminal of the first arithmetic unit; the output terminal of the controlled object is connected to the input terminal of the attitude feedback module, and the output terminal of the attitude feedback module is connected to the third input terminal of the first arithmetic unit.

[0021] The formula for calculating the total transfer function G(s) of the servo control system is as follows:

[0022]

[0023] Wherein, C(s) is the output function of the AGV servo control system, and R(s) is the input function of the AGV servo control system; G1(s) and G2(s) are abstract functions of the servo controller in the servo control system, G3(s) is an abstract function of the servo driver in the servo control system, G4(s) is an abstract function of the servo motor in the servo control system, G5(s) is an abstract function of the speed reduction device in the servo control system; H2(s) is an abstract function of the speed feedback module in the servo control system, and H3(s) is an abstract function of the current feedback module in the servo control system.

[0024] Furthermore, step 3 specifically involves:

[0025] Step 3.1: Initialize the neural network, which includes an input layer, a hidden layer, and an output layer arranged sequentially. Set the input layer weight set {ω}. it}, Hidden layer weight set {ω ij} and the set of output layer weights {ω ik}, and the input layer threshold set {θ t}, Hidden layer threshold set {θ j} and the 00 output layer threshold set {θ k}, define the number of learning iterations N = 0, the preset error e, and the maximum number of learning iterations N1; where i, j, t, and k represent the position information of neurons in different layers, the first letter of the ω subscript indicates the number of layers in the neural network, and the second letter of the ω subscript indicates the sequence number of the neuron; each layer of the neural network has one and only one threshold, which represents the connection relationship between different layers;

[0026] Step 3.2: Calculate the input and output values ​​of the hidden layer neurons; the activation function of the hidden layer is... Where y j x is the output value of the hidden layer. j Here, m is the input variable for the hidden layer, and m is the number of records.

[0027] Step 3.3: Based on the observation equation Z of the AGV relative to the target point and the total transfer function G(s) of the servo control system, calculate the input and output values ​​of the output layer neurons; the activation function of the output layer is expressed as... Where y k g is the output value of the output layer. k For the input variables of the output layer; f(x) = tanx, β is an integer;

[0028] Step 3.4: Calculate the generalized error of each neuron in the hidden layer and the output layer. and Then let N = N + 1;

[0029] and The specific calculation formula is as follows:

[0030]

[0031] Among them, is the input variable of the output layer neuron, is the input variable of the input layer neuron, is the input variable of the hidden layer neuron;

[0032]

[0033] is the expected output value, is the actual output value; is the output error value,

[0034] z is the data recording variable, ω zk 、ω zt 、ω zj and s z are all input variables; a, b, and n are all the number of recording times, and their values are automatically taken by the system;

[0035] Step 3.5, judge the value of N. If N≥N1, then execute Step 3.7. If N<N1, then compare and with e respectively. If and are both less than e, then execute Step 3.7. If and at least one is greater than or equal to e, then execute Step 3.6;

[0036] For Step 3.6, use and obtained in Step 3.4 to update the weights ω it 、ω ij and ω tk , as well as the thresholds θ t 、θ j and θ k , and then return to Step 3.2; the specific calculation formula is as follows:

[0037]

[0038] Among them, η and α are both learning rates, 0<η<1, 0<α<1; p is the number of recording times, and its value is automatically taken by the system; Δω is the change value of the weights between two consecutive times, and △θ is the change value of the thresholds between two consecutive times; and These are the input variables for neurons in the hidden layer, output layer, and input layer, respectively.

[0039] Step 3.7: Obtain the neural network training result, i.e., the optimized input layer weight set {ω}. it}, Hidden layer weight set {ω ij}、 Output layer weight set {ω tk}, Input layer threshold set {θ t}, Hidden layer threshold set {θ j} and the output layer threshold set {θ k}

[0040] Meanwhile, the present invention also provides a computer storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the steps of the above-mentioned AGV path tracking and optimization method based on a DSP neural network algorithm.

[0041] In addition, the present invention also provides a computer device, including a processor, a memory connected to the processor, and a computer program that can run on the processor, wherein the processor executes the computer program to implement the steps of the above-mentioned AGV path tracking and optimization method based on a DSP neural network algorithm.

[0042] Compared with the prior art, the beneficial effects of the present invention are:

[0043] The AGV path tracking and optimization method based on DSP neural network algorithm provided by this invention utilizes neural network for sensor data processing and fusion. By integrating the observations provided by different types of sensors through data fusion, it reduces potential contradictions between multiple sensors, reduces environmental interference, and thus improves the accuracy of equipment status detection system faults. At the same time, based on the advantages of neural network nonlinear mapping capability, self-learning capability, generalization capability, and fault tolerance capability, it can enhance the stability of the system and improve the overall system response and processing speed. Attached Figure Description

[0044] Figure 1 This is a flowchart illustrating an embodiment of the AGV path tracking and optimization method based on a DSP-based neural network algorithm according to the present invention.

[0045] Figure 2 This is a flowchart of step 3 in an embodiment of the AGV path tracking and optimization method based on DSP neural network algorithm of the present invention;

[0046] Figure 3 The results of the MATLAB simulation test in step 4 of the embodiment of the AGV path tracking and optimization method based on DSP neural network algorithm of the present invention are as follows. Figure 1 ;

[0047] Figure 4 The results of the MATLAB simulation test in step 4 of the embodiment of the AGV path tracking and optimization method based on DSP neural network algorithm of the present invention are as follows. Figure 2 ;

[0048] Figure 5 This is a schematic diagram of the mathematical model of the servo control system of the AGV in step 2 of an embodiment of the AGV path tracking and optimization method based on DSP neural network algorithm of the present invention.

[0049] Figure 6 This is a schematic diagram of the neural network structure constructed in step 3 of an embodiment of the AGV path tracking and optimization method based on DSP neural network algorithm of the present invention;

[0050] Figure 7 This is a schematic diagram of the control circuit structure of an AGV;

[0051] Explanation of reference numerals in the attached figures:

[0052] 21. Servo Controller; 22. Servo Driver; 23. Servo Motor; 24. Speed ​​Reducer; 25. Controlled Object; 26. Attitude Feedback Module; 27. Speed ​​Feedback Module; 28. Current Feedback Module; 31. Input Layer; 32. Hidden Layer; 33. Output Layer; 61. Environmental Data; 62. Sensor; 63. Amplifier Circuit; 64. Switch Signal; 65. Other Input Signals; 66. DSPTMS320F2812 Module; 6601. DI Module; 6602. ADC Module; 6603. Driver Software; 6604. eCAN Module; 6605. Other Input Modules; 67. JTAG Interface Circuit; 68. Clock Oscillation; 69. Power Supply Circuit; 610. Host Computer. Detailed Implementation

[0053] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0054] A method for AGV path tracking and optimization based on a DSP-based neural network algorithm, see [link / reference]. Figure 1 This includes the following steps:

[0055] Step 1: Establish a mathematical model of the AGV's motion state, obtain the motion state equations of the AGV in the two-dimensional plane at time t and at time t+Δt, and calculate the observation equation Z of the AGV relative to the target point based on the motion state equations of the AGV in the two-dimensional plane at time t and at time t+Δt.

[0056] The motion state equation of the AGV in the two-dimensional plane at time t is:

[0057]

[0058] Where t is time, xt and yt are the x-direction and y-direction positions of the AGV at time t, ut is the angle between the AGV's motion direction at time t and the x-axis, vt is the velocity of the AGV at time t, ωt is the angular velocity of the AGV at time t, and at is the acceleration of the AGV at time t.

[0059] d is the differential symbol;

[0060] At time t+Δt, the motion state equation of the AGV in the two-dimensional plane is:

[0061]

[0062] Among them, V max and V min These represent the maximum and minimum speeds of the AGV between time t and t+Δt, respectively.

[0063] At time t, the observation equation Z of the AGV relative to the target point is:

[0064]

[0065] Where D represents the distance between the current position of the AGV and the target point. x is the azimuth angle of the AGV's current position relative to the target point. d and y d Let x and y represent the position coordinates of the target point, respectively. t and y t Let x and y represent the position coordinates of the AGV at time t, respectively.

[0066] Step 2: Establish a mathematical model of the AGV's servo control system and obtain the overall transfer function Gs of the servo control system; see [link to servo control system mathematical model] for details. Figure 5The system includes a servo driver 22, a servo motor 23, a speed reducer 24, and a controlled object 25 connected in sequence, as well as a servo controller 21, a current feedback module 28, a speed feedback module 27, an attitude feedback module 26, a first arithmetic unit, and a second arithmetic unit. The first input terminal of the first arithmetic unit is connected to the input terminal of the AGV servo control system, and the output terminal is connected to the input terminal of the servo controller 21. The output terminal of the servo controller 21 is connected to the first input terminal of the second arithmetic unit. The input terminal of the current feedback module 28 is connected between the servo motor 23 and the speed reducer 24, and the output terminal is connected to the second input terminal of the second arithmetic unit. The output terminal of the second arithmetic unit is connected to the input terminal of the servo driver 22. The input terminal of the speed feedback module 27 is located between the speed reducer 24 and the controlled object 25, and the output terminal is connected to the second input terminal of the first arithmetic unit. The output terminal of the controlled object 25 is connected to the input terminal of the attitude feedback module 26, and the output terminal of the attitude feedback module 26 is connected to the third input terminal of the first arithmetic unit.

[0067] The formula for calculating the total transfer function G(s) of a servo control system is:

[0068]

[0069] Wherein, C(s) is the output function of the AGV servo control system, and R(s) is the input function of the AGV servo control system; G1(s) and G2(s) are abstract functions of the servo controller 21 in the servo control system, G3(s) is an abstract function of the servo driver 22 in the servo control system, G4(s) is an abstract function of the servo motor 23 in the servo control system, G5(s) is an abstract function of the speed reduction device 24 in the servo control system; H2(s) is an abstract function of the speed feedback module 27 in the servo control system, and H3(s) is an abstract function of the current feedback module 28 in the servo control system.

[0070] Step 3, see Figure 2 Based on the observation equation Z of the AGV relative to the target point and the total transfer function G(s) of the servo control system, a neural network is constructed and learned until the error reaches the preset requirement or the number of learning times reaches the preset requirement, and the neural network training result is obtained.

[0071] Step 3.1: Initialize the neural network, see [link / reference] Figure 6 The neural network includes an input layer 31, a hidden layer 32, and an output layer 33 arranged sequentially, and the input layer weight set {ω} is defined. it}, Hidden layer weight set {ω ij} and the set of output layer weights {ω ik}, and the input layer threshold set {θ t}, Hidden layer threshold set {θ j} and the 00 output layer threshold set {θk}, define the number of learning iterations N = 0, the preset error e, and the maximum number of learning iterations N1; where i, j, t, and k represent the position information of neurons in different layers, the first letter of the ω subscript indicates the number of layers in the neural network, and the second letter of the ω subscript indicates the sequence number of the neuron; each layer of the neural network has one and only one threshold, which represents the connection relationship between different layers;

[0072] Step 3.2: Calculate the input and output values ​​of neurons in hidden layer 32; the activation function of hidden layer 32 is... Where y j x is the output value of hidden layer 32. j Here, m is the input variable for hidden layer 32, and m is the number of records.

[0073] Step 3.3: Based on the observation equation Z of the AGV relative to the target point and the total transfer function Gs of the servo control system, calculate the input and output values ​​of neurons in output layer 33; the activation function of output layer 33 is expressed as... Where y k For the output value of output layer 33, g k For the input variables of output layer 33; f(x) = tanx, β is an integer;

[0074] Step 3.4: Calculate the generalized error of each neuron in hidden layer 32 and output layer 33. and Then let N = N + 1;

[0075] and The specific calculation formula is as follows:

[0076]

[0077] in, For the input variables of neurons in the output layer 33, For the input variables of neurons in layer 31, For the input variables of the 32 neurons in the hidden layer;

[0078]

[0079] For the expected output value, This is the actual output value; To output the error value,

[0080] z is a data record variable, ω zk ω zt ω zj and s zAll are input variables; a, b, and n are all the number of records, and their values are automatically assigned by the system;

[0081] Step 3.5: Judge the value of N. If N ≥ N1, execute Step 3.7. If N < N1, then and are respectively compared with e. If and are both less than e, execute Step 3.7. If and at least one is greater than or equal to e, execute Step 3.6;

[0082] Step 3.6: Use and obtained in Step 3.4 to update the weight ω it , ω ij and ω tk , as well as the threshold values θ t , θ j and θ k , and then return to Step 3.2; the specific calculation formula is:

[0083]

[0084] where η and α are both learning rates, 0 < η < 1, 0 < α < 1; p is the number of records, and its value is automatically assigned by the system; Δω is the change value of the weight between two consecutive times, and △θ is the change value of the threshold between two consecutive times; and are respectively the input variables of the neurons in the 32nd hidden layer, 33rd output layer, and 31st input layer;

[0085] Step 3.7: Obtain the neural network training results, that is, the optimized input layer weight set {ω it}, hidden layer weight set {ω ij}, output layer weight set {ω tk}, input layer threshold set {θ t}, hidden layer threshold set {θ j}, and output layer threshold set {θ k}.

[0086] Step 4: In the computer software, call the neural network training results obtained in Step 3 to optimize the observation equation Z of the AGV relative to the target point and the total transfer function G(s) of the servo control system, to obtain the optimized observation equation Z1 of the AGV relative to the target point and the optimized total transfer function G1(s) of the servo control system. Then, call the neural network training results obtained in Step 3, and perform simulation tests based on Z1 and G1(s) to obtain the test results. If the test results do not meet the preset requirements, update Z and G(s) using Z1 and G1(s), return to Step 3, and rebuild the neural network for training. If the test results meet the preset requirements, the path planning of the AGV is completed.

[0087] The test results for this embodiment can be found in [link / reference]. Figure 3 and Figure 4 ,in Figure 3 The x-axis represents the number of test sets, and the y-axis represents the simulation output results. Figure 3 The blue curve represents the actual output curve, and the orange curve represents the simulated output curve. Figure 3 It can be seen that the fitting effect is good. Figure 4 The horizontal axis represents the step size, and the vertical axis represents the mean variance. Figure 4 The blue curve represents the training output error curve, the green curve represents the actual output error curve, and the red curve represents the test output error curve. Figure 4 It can be seen that when the test step size is 35, the set error can be achieved, which can meet the timeliness requirements of actual engineering.

[0088] The TMS series DSPs feature high processing speed and a system architecture and instructions suitable for digital signal processing. The DSP chip used is the TI TMS320F2812, a high-performance floating-point digital signal processor with a Harvard architecture, multiple adders and multipliers, 256KB of on-chip memory, and an operating frequency of 200MHz. Since the implementation of neural networks involves numerous matrix inner products and multiply-accumulate operations, it is difficult and slow to implement using ordinary microcontrollers. Applying the obtained AGV path planning algorithm to the DSP hardware platform can fully utilize the high-speed processing capabilities of the DSP, enabling online data processing of rotating machinery. Figure 7 The hardware circuit system utilizes the rich peripheral modules of the TMS320F2812 to connect external signal processing and communication circuits, forming a hardware platform. The hardware circuit mainly includes a data acquisition module, a digital input / output module, and a CAN bus communication module. The specific control software can be written using CCS compiler software. See the schematic diagram of the AGV control circuit structure. Figure 7 In practical applications, after completing the algorithm design, it is also necessary to adapt the control circuit of the AGV.

[0089] In practical applications, when an AGV receives operating instructions and coordinate parameters, it acquires position and environmental information through its own sensors. Combined with data from the dispatch center or host computer, the AGV autonomously plans its pre-route based on the "shortest distance principle" and begins operation. During task execution, the AGV receives data from the dispatch center or host computer in real time and acquires environmental information through sensors, obtaining measurement data. After nonlinear mapping by a neural network, the data from each sensor are independent. Then, the data from each independent sensor is fused together using a global neural network. When noise is present, using multiple sensors for identification and then fusing the results from each sensor improves the recognition rate. Simultaneously, neural network learning methods are used for perception and observation, resulting in specific state feature representations. The neural network algorithm enables the AGV to directly output control from environmental perception input to action, helping it complete autonomous navigation and obstacle avoidance tasks. The AGV continuously adjusts and plans its path during operation to complete its task.

[0090] In addition, this embodiment also provides a computer storage medium on which a computer program is stored. When the computer program is executed by a processor, it implements the steps of the above-mentioned AGV path tracking and optimization method based on a DSP neural network algorithm.

[0091] In addition, this embodiment also provides a computer device, including a processor, a memory connected to the processor, and a computer program that can run on the processor. When the processor executes the computer program, it implements the steps of the above-described AGV path tracking and optimization method based on a DSP neural network algorithm.

Claims

1. A method for AGV path tracking and optimization based on a DSP-based neural network algorithm, characterized in that, Includes the following steps: Step 1: Establish a mathematical model of the AGV's motion state, obtain the motion state equations of the AGV in the two-dimensional plane at time t and at time t+Δt, and calculate the observation equation Z of the AGV relative to the target point based on the motion state equations of the AGV in the two-dimensional plane at time t and at time t+Δt. Step 2: Establish a mathematical model of the servo control system of the AGV and obtain the total transfer function G(s) of the servo control system; Step 3: Based on the observation equation Z of the AGV relative to the target point and the total transfer function G(s) of the servo control system, construct a neural network and learn it until the error reaches the preset requirement or the number of learning iterations reaches the preset requirement, and obtain the neural network training results. Step 3.1: Initialize the neural network, which includes an input layer (31), a hidden layer (32), and an output layer (33) arranged sequentially. The input layer weight set {ω} is set. it }, Hidden layer weight set {ω ij } and the set of output layer weights {ω ik }, and the input layer threshold set {θ t }, Hidden layer threshold set {θ j } and the 00 output layer threshold set {θ k }, define the number of learning iterations N = 0, the preset error e, and the maximum number of learning iterations N1; where i, j, t, and k represent the position information of neurons in different layers, the first letter of the ω subscript indicates the number of layers in the neural network, and the second letter of the ω subscript indicates the sequence number of the neuron; each layer of the neural network has one and only one threshold, which represents the connection relationship between different layers; Step 3.2: Calculate the input and output values ​​of the neurons in the hidden layer (32); the activation function of the hidden layer (32) is... Where y j x is the output value of hidden layer (32). j is the input variable for the hidden layer (32), and m is the number of records; Step 3.3: Based on the observation equation Z of the AGV relative to the target point and the total transfer function G(s) of the servo control system, calculate the input and output values ​​of the neurons in the output layer (33); the activation function of the output layer (33) is expressed as... Where y k For the output value of the output layer (33), g k The input variables for the output layer (33); f(x) = tanx, β is an integer; Step 3.4: Calculate the generalized error of each neuron in the hidden layer (32) and the output layer (33). and Then let N = N + 1; and The specific calculation formula is as follows: in, For the input variables of the output layer (33) neurons, For the input variables of the neurons in the input layer (31), For the input variables of the hidden layer (32) neurons; For the expected output value, This is the actual output value; To output the error value, z is a data record variable, ω zk ω zt ω zj and s z All are input variables; a, b, and n are the number of records, and their values ​​are automatically assigned by the system. Step 3.5: Determine the value of N. If N ≥ N1, then execute Step 3.

7. If N < N1, then compare and with e respectively. If and are both less than e, then execute Step 3.

7. If and at least one is greater than or equal to e, then execute Step 3.6; Step 3.6: Using the results obtained in step 3.4 and For weight ω it ω ij and ω tk and threshold θ t θ j and θ k Perform the update, then return to step 3.2; the specific calculation formula is as follows: Where η and α are both learning rates, 0 < η < 1, 0 < α < 1; p is the number of records, which is automatically assigned by the system; Δω is the change in weight between two consecutive records, and Δθ is the change in threshold between two consecutive records. and These are the input variables of neurons in the hidden layer (32), output layer (33), and input layer (31), respectively; Step 3.7: Obtain the neural network training result, i.e., the optimized input layer weight set {ω}. it }, Hidden layer weight set {ω ij }、 Output layer weight set {ω tk }, Input layer threshold set {θ t }, Hidden layer threshold set {θ j } and the output layer threshold set {θ k }; Step 4: In the computer software, call the neural network training results obtained in Step 3 to optimize the observation equation Z of the AGV relative to the target point and the total transfer function G(s) of the servo control system, to obtain the optimized observation equation Z1 of the AGV relative to the target point and the optimized total transfer function G1(s) of the servo control system. Then, call the neural network training results obtained in Step 3, and perform simulation tests based on Z1 and G1(s) to obtain the test results. If the test results do not meet the preset requirements, update Z and G(s) using Z1 and G1(s), return to Step 3, and rebuild the neural network for training. If the test results meet the preset requirements, the path planning of the AGV is completed.

2. The AGV path tracking and optimization method based on a DSP-based neural network algorithm according to claim 1, characterized in that: In step 1, the motion state equation of the AGV in the two-dimensional plane at time t is: Where t is time, x(t) and y(t) are the x-direction and y-direction positions of the AGV at time t, respectively, u(t) is the angle between the AGV's motion direction at time t and the x-axis, v(t) is the velocity of the AGV at time t, ω(t) is the angular velocity of the AGV at time t, and a(t) is the acceleration of the AGV at time t. d is the differential symbol; At time t+Δt, the motion state equation of the AGV in the two-dimensional plane is: Among them, V max and V min These represent the maximum and minimum speeds of the AGV between time t and t+Δt, respectively. At time t, the observation equation Z of the AGV relative to the target point is: Where D represents the distance between the current position of the AGV and the target point. x is the azimuth angle of the AGV's current position relative to the target point. d and y d Let x and y represent the position coordinates of the target point, respectively. t and y t Let x and y represent the position coordinates of the AGV at time t, respectively.

3. The AGV path tracking and optimization method based on a DSP-based neural network algorithm according to claim 2, characterized in that: In step 2, the mathematical model of the servo control system includes a servo driver (22), a servo motor (23), a speed reduction device (24), and a controlled object (25) connected in sequence, as well as a servo controller (21), a current feedback module (28), a speed feedback module (27), an attitude feedback module (26), a first arithmetic unit, and a second arithmetic unit; the first input terminal of the first arithmetic unit is connected to the input terminal of the AGV servo control system, and the output terminal is connected to the input terminal of the servo controller (21). The output terminal of the servo controller (21) is connected to the first input terminal of the second arithmetic unit. The input of the flow feedback module (28) is connected between the servo motor (23) and the speed reducer (24), and the output is connected to the second input of the second arithmetic unit. The output of the second arithmetic unit is connected to the input of the servo driver (22). The input of the speed feedback module (27) is located between the speed reducer (24) and the controlled object (25), and the output is connected to the second input of the first arithmetic unit. The output of the controlled object (25) is connected to the input of the attitude feedback module (26), and the output of the attitude feedback module (26) is connected to the third input of the first arithmetic unit. The formula for calculating the total transfer function G(s) of the servo control system is as follows: Wherein, C(s) is the output function of the AGV servo control system, and R(s) is the input function of the AGV servo control system; G1(s) and G2(s) are abstract functions of the servo controller (21) in the servo control system, G3(s) is an abstract function of the servo driver (22) in the servo control system, G4(s) is an abstract function of the servo motor (23) in the servo control system, and G5(s) is an abstract function of the speed reduction device (24) in the servo control system; H2(s) is an abstract function of the speed feedback module (27) in the servo control system, and H3(s) is an abstract function of the current feedback module (28) in the servo control system.

4. A computer storage medium storing a computer program thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the AGV path tracking and optimization method based on a DSP neural network algorithm as described in any one of claims 1-3.

5. A computer device comprising a processor, a memory connected to the processor, and a computer program executable on the processor, characterized in that: When the processor executes the computer program, it implements the steps of the AGV path tracking and optimization method based on a DSP neural network algorithm as described in any one of claims 1-3.

Images