Air spring adjustable seat suspension control method, system, seat, and medium

Abstract

The present application relates to air spring adjustable seat suspension control method, system, seat and medium, including: step 1. Respectively obtain the time sequence speed signal of the seat suspension upper bottom plate in the air spring adjustable seat suspension and the elastic force of the air spring in the air spring adjustable seat suspension;Step 2. The time sequence speed signal collected is handled and the speed of the seat suspension upper bottom plate is predicted by using RBF radial basis neural network prediction model;Step 3. The predicted speed of the seat suspension upper bottom plate is used as the input of Fuzzy-PID control algorithm, and the corresponding control force of the air spring is obtained;Step 4. When the predicted control force of the air spring is greater than or equal to the elastic force of the air spring, open the switch of the inflation electromagnetic valve, and inflate the air spring, when the predicted control force of the air spring is less than the elastic force of the air spring, open the switch of the exhaust electromagnetic valve.The present application improves the negative influence caused by the response time lag of the air spring.

Description

Technical Field

[0001] This invention relates to the field of air spring adjustable seat suspension control technology, and in particular to an air spring adjustable seat suspension control method, system, seat, and medium. Background Technology

[0002] With the rapid development and widespread use of transportation tools such as automobiles, airplanes, and trains, people have higher expectations for comfort during travel. Traditional seat suspension systems have certain limitations in providing support and adaptability. Against this backdrop, researchers are working to provide drivers with a more intelligent and responsive suspension system by integrating advanced air spring technology, thereby improving the ride comfort of automotive seat suspensions.

[0003] Semi-active seat suspensions have the ability to adjust damper damping or spring stiffness within a certain range, and their system requires relatively little energy. Air-spring adjustable semi-active seat suspensions are a typical example of stiffness-adjustable semi-active seat suspensions, which adjust stiffness by regulating the airflow through the air springs. Therefore, during adjustment, response delay and overcharging / over-discharging can lead to deterioration of control performance and excessively high or low air pressure, thus affecting seat stability and comfort. Therefore, although this adjustment method has certain advantages in terms of control and cost, issues such as response delay and overcharging / over-discharging need to be carefully addressed to ensure the performance and reliability of the seat suspension system.

[0004] Currently, various control strategies have been proposed for adjustable air spring seat suspension control technology. However, when studying control strategies for the suspension system itself, the actuator response problem is often overlooked. For adjustable air springs, the inflation and deflation process requires a certain amount of time. The larger the air spring volume, the longer the inflation time, and the longer the response time of the actuator itself. If these issues are ignored when designing control strategies, phenomena such as deteriorated control performance and over-inflation and over-discharge may occur.

[0005] Therefore, it is necessary to develop an air-spring adjustable seat suspension control method, system, seat, and medium. Summary of the Invention

[0006] The purpose of this invention is to provide an air spring adjustable seat suspension control method, system, seat and medium, which can improve the negative impact caused by the response lag of the air spring, thereby improving the vibration isolation effect of the seat suspension and the riding comfort of the driver and passengers.

[0007] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0008] In a first aspect, the present invention provides an air spring adjustable seat suspension control method, comprising the following steps:

[0009] Step 1. Obtain the time-series velocity signal of the upper plate of the seat suspension in the adjustable air spring seat suspension and the elastic force of the air spring in the adjustable air spring seat suspension, respectively;

[0010] Step 2. Use the RBF radial basis function neural network prediction model to process the collected time series velocity signals and predict the velocity of the upper plate of the seat suspension;

[0011] Step 3. Use the predicted speed of the seat suspension upper plate as the input to the Fuzzy-PID control algorithm to obtain the corresponding control force of the air spring;

[0012] Step 4. When the predicted control force of the air spring is greater than or equal to the spring force of the air spring, open the air inflation solenoid valve to inflate the air spring. When the predicted control force of the air spring is less than the spring force of the air spring, open the air release solenoid valve to change the spring force of the air spring.

[0013] Optionally, the method for constructing the RBF radial basis function neural network prediction model in step 2 is as follows:

[0014] Set the number of nodes in the input, hidden, and output layers of the RBF neural network; set the time series delay step size of the RBF neural network to d and the prediction step size to p;

[0015] The activation function of the hidden layer is set to a Gaussian function, as shown in the following formula:

[0016]

[0017] Where: x i c represents the input corresponding to the i-th hidden layer node in the input sample; i δ represents the center of the basis function of the i-th hidden layer node; i This represents the expansion constant of the i-th hidden layer node;

[0018] After the data passes through the activation function in the hidden layer, the hidden layer output is as follows:

[0019]

[0020] Where: φ(x) i ,c i ) represents the output of the i-th hidden layer node;

[0021] A linear mapping is achieved from the hidden layer to the output layer, as shown in the following formula:

[0022]

[0023] Where: y represents the output of the output layer; j represents the number of hidden layer nodes; w i This represents the weight of the i-th hidden layer node;

[0024] The parameters of the RBF neural network prediction model are corrected based on the training set data; and when the parameters of the RBF neural network prediction model meet the usage requirements, the optimal parameter set of the RBF neural network prediction model is determined, thus obtaining the trained RBF neural network prediction model.

[0025] Optionally, the number of input layers is 7; the number of nodes in the hidden layers is 7; and the number of nodes in the output layers is 1.

[0026] Optionally, if the prediction error is within a set threshold and has not reached a preset number of iterations, it indicates that the parameters of the RBF neural network prediction model meet the requirements; otherwise, it indicates that the parameters of the RBF neural network prediction model do not meet the requirements, and the parameters of the RBF neural network prediction model are corrected using the gradient descent method; the specific formula is as follows:

[0027]

[0028]

[0029]

[0030] Where α represents the learning rate; k represents the number of iterations; and D represents the error function.

[0031] Optionally, step 3 specifically includes the following steps:

[0032] Step 31. Transmit the speed signal of the seat suspension floor plate predicted by the RBF neural network prediction model to the Fuzzy-PID control module;

[0033] Step 32. Set the desired seat suspension speed signal for the Fuzzy-PID control module and transmit this signal to the Fuzzy-PID control module;

[0034] Step 33. The Fuzzy-PID control module calculates the deviation using the signals obtained in steps 31 and 32, and outputs the control force of the air spring using the Fuzzy-PID control algorithm.

[0035] Optionally, step 33 specifically includes the following steps:

[0036] Step 331. Compare the predicted speed signal of the seat suspension upper plate with the desired seat suspension speed signal, calculate the error signal e and its error change rate ec, and use them as input variables for the Fuzzy-PID control module. The output fuzzy control variable is ΔK. p ΔK i ΔK d ;

[0037] Step 332. The error signal e and its rate of change ec are used as input variables of the Fuzzy-PID control module. After the quantization factor is applied, the corresponding fuzzy variables E and EC are obtained, and the basic physical domain is transformed into a fuzzy domain. Specifically, the fuzzy subsets are taken as {NB, NM, NS, O, PS, PM, PB}, and the fuzzy terms are divided into 7 levels, namely -3, -2, -1, 0, 1, 2, 3. The membership function of each fuzzy subset is selected as a Gaussian membership function to describe it. Among them, NB represents negative large, NM represents negative medium, NS represents negative small, PS represents positive small, PM represents positive medium, and PB represents positive large.

[0038] Step 333. Perform fuzzy inference using fuzzy rules, that is, look up the fuzzy inference table based on the membership degrees of the two input signals, the error signal e and its rate of change of error ec, to obtain the output fuzzy control vector ΔK. p ΔK i ΔK d ;

[0039] Step 334. Apply the centroid method to the output fuzzy control vector ΔK. p ΔK i ΔK d After defuzzing, the parameters of the traditional PID controller are adjusted to achieve adaptive adjustment of the PID's proportional, integral, and derivative coefficients.

[0040] Secondly, the present invention provides an air spring adjustable seat suspension control system, comprising a controller and a memory, wherein the memory stores a computer-readable program, and the computer-readable program, when invoked by the controller, can execute the steps of the air spring adjustable seat suspension control method as described in the present invention.

[0041] Thirdly, the seat described in this invention employs an air spring adjustable seat suspension control system as described in this invention.

[0042] Fourthly, the present invention provides a medium storing a computer-readable program that, when invoked, can execute the steps of the air spring adjustable seat suspension control method described in the present invention.

[0043] The present invention has the following beneficial effects:

[0044] (1) Considering that response delay and overcharging or over-discharging may lead to deterioration of control effect and excessively high or low gas pressure, the present invention uses the RBF radial basis neural network prediction model to process the collected data information and predict the speed of the seat suspension floor plate. The predicted speed signal is used as the input of the Fuzzy-PID control module to predict the corresponding control force of the air spring, thereby opening the control solenoid valve for inflation or deflation in advance, and inflating or deflating the air spring in advance, thereby changing the elastic force of the air spring and minimizing the impact of the air spring inflation and deflation time delay.

[0045] (2) This invention overcomes the shortcomings of traditional PID controllers by designing a rule base for fuzzy controllers and adaptively adjusting the proportional, integral and derivative coefficients of the PID controller according to different speed signals, so that the air spring adjustable seat suspension can achieve precise control, thereby improving the vibration isolation performance of the air spring adjustable seat suspension and the riding comfort of the driver and passengers. Attached Figure Description

[0046] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 This is a flowchart of the air spring adjustable seat suspension control method described in the embodiments of this application;

[0048] Figure 2 This is a flowchart illustrating the construction of the RBF radial basis neural network prediction model described in the embodiments of this application;

[0049] Figure 3 This is a schematic diagram of the structure of the RBF radial basis neural network prediction model described in the embodiments of this application;

[0050] Figure 4 This is a schematic diagram of the principle of the Fuzzy-PID control model described in the embodiments of this application;

[0051] Figure 5 This is a schematic diagram of the air spring adjustable seat suspension control system described in the embodiments of this application;

[0052] Figure 6 This is a schematic diagram of the air spring adjustable seat suspension control system described in the embodiments of this application. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0054] To enable those skilled in the art to better understand the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0055] First, we will introduce the air-spring adjustable seat suspension, a current technology. It generally includes an air spring (using an adjustable air spring), an inflation solenoid valve, a deflation solenoid valve, an air compressor, gas lines, damping shock absorbers, a seat transmission mechanism, and a seat transmission bracket. The air compressor, gas lines, inflation solenoid valve, deflation solenoid valve, and air spring constitute the adjustable module of the seat suspension system. The seat transmission bracket includes an upper seat suspension plate and a lower seat suspension plate. The seat transmission mechanism is "X"-shaped, with its upper end connected to the upper seat suspension plate and its lower end connected to the lower seat suspension plate. The damping shock absorber, as a passive component, primarily provides resistance to motion, absorbing or reducing vibration energy as quickly as possible. During the relative movement of the upper and lower seat suspension plates, the seat transmission mechanism provides motion guidance and load bearing for the seat transmission bracket. The upper end of the air spring is connected to the upper base plate of the seat suspension, and the lower end of the air spring is connected to the lower base plate of the seat suspension. The air spring is connected to the inflation solenoid valve and the deflation solenoid valve through pneumatic lines. The inflation solenoid valve and the deflation solenoid valve are also connected to the air compressor through pneumatic lines. The inflation solenoid valve and the deflation solenoid valve are also connected to the controller, which controls the opening and closing of the inflation solenoid valve and the deflation solenoid valve, thereby adjusting the amount of gas in the air spring airbag and ultimately adjusting the elasticity of the air spring.

[0056] In this embodiment, a speed sensor and two displacement sensors are also required. The speed sensor is positioned on the upper base plate of the seat suspension, and the two displacement sensors are respectively positioned on the upper and lower base plates of the seat suspension. The speed sensor measures the speed signal of the upper base plate, and the two displacement sensors measure the displacement signals of the upper and lower base plates. Both the speed sensor and the displacement sensors are connected to a controller. The speed sensor acquires the time-series speed signal of the upper base plate, and the displacement sensors acquire the displacement signals of the upper and lower base plates.

[0057] like Figure 1 and Figure 5 As shown in the embodiments of this application, a method for controlling an air spring adjustable seat suspension includes:

[0058] Step 1. Obtain the time-series velocity signal of the upper plate of the seat suspension in the adjustable air spring seat suspension and the elastic force of the air spring in the adjustable air spring seat suspension, respectively;

[0059] Step 2. Use the RBF radial basis function neural network prediction model to process the collected time series velocity signals and predict the velocity of the upper plate of the seat suspension;

[0060] Step 3. Use the predicted speed of the seat suspension upper plate as the input to the Fuzzy-PID control algorithm to obtain the corresponding control force of the air spring;

[0061] Step 4. When the predicted control force of the air spring is greater than or equal to the spring force, open the inflation solenoid valve to inflate the air spring. When the predicted control force of the air spring is less than the spring force, open the deflation solenoid valve. This allows for pre-emptive inflation or deflation of the air spring to change its spring force.

[0062] In one possible embodiment, step 1 specifically involves: acquiring a time-series speed signal of the seat suspension's floor plate using a speed sensor; and calculating the air spring's spring force using data collected by a pressure sensor integrated within the air spring.

[0063] like Figure 2 and Figure 3 As shown, in one possible embodiment, the method for constructing the RBF radial basis function neural network prediction model is as follows:

[0064] (1) Determine the structure of the RBF neural network:

[0065] Based on the three-layer structure of the RBF neural network (input layer – hidden layer – output layer) and considering the characteristics of the target object, after extensive experimental design and performance comparison, a 7-7-1 RBF neuron structure was selected, consisting of 7 input layers, 7 hidden layer nodes, and 1 output layer node. Since only the velocity signal value of the seat suspension's floor plate needs to be predicted, only 1 output layer node was chosen. The number of hidden layer nodes and input layer nodes was determined through extensive debugging and experimental design. For example, a higher number of input layers theoretically leads to more accurate predictions, but it also reduces computational training efficiency and can cause overfitting.

[0066] (2) RBF neural network parameter configuration:

[0067] By setting a time series delay step size d and a prediction step size p, a special method for delay step size is proposed (for example, this function can be implemented using delay module groups in Simulink). This enables the RBF radial basis neural network prediction model to predict the speed signal of the seat suspension floor plate at future times (e.g., t+1, t+2, ..., t+p) in real time based on the most recent data information (e.g., the most recent data information includes t, t-1, t-2, ..., td). This achieves the goal of the RBF radial neural network predicting the speed signal of the seat suspension floor plate at future times (e.g., 2 steps) in real time using historical data information of d steps (e.g., 7 steps).

[0068] The activation function of the hidden layer is set to a Gaussian function, as shown in the following formula:

[0069]

[0070] Where: x i c represents the input corresponding to the i-th hidden layer node in the input sample; i δ represents the center of the basis function of the i-th hidden layer node; i This represents the expansion constant of the i-th hidden layer node;

[0071] When the data is in the hidden layer, after passing through the activation function, the hidden layer output is as follows:

[0072]

[0073] Where: φ(x) i ,c i ) represents the output of the i-th hidden layer node.

[0074] A linear mapping is achieved from the hidden layer to the output layer, as shown in the following formula:

[0075]

[0076] Where: y represents the output of the output layer; j represents the number of hidden layer nodes; w i This represents the weight of the i-th hidden layer node;

[0077] (3) Model training:

[0078] The parameters of the RBF radial basis function neural network prediction model are corrected based on the training set data. If the prediction error is within the error threshold and has not reached the preset number of iterations, the parameters of the RBF radial basis function neural network prediction model meet the usage requirements. Otherwise, the parameters of the RBF radial basis function neural network prediction model are corrected using the gradient descent method. The specific formula is as follows:

[0079]

[0080]

[0081]

[0082] Where α represents the learning rate; k represents the number of iterations; and D represents the error function.

[0083] Once the optimal parameter set is determined, the RBF radial basis function neural network prediction model can be established. After initialization and model definition, the RBF radial basis function neural network prediction model can be trained. After training, the required data can be used to predict the speed signal value of the seat suspension floor plate at future times.

[0084] In one possible embodiment, step 2 specifically involves:

[0085] Obtain the time series velocity signal value corresponding to the number of input layers of the RBF radial basis neural network prediction model, that is, each prediction uses historical data information with a set step size to predict future data;

[0086] The time-series velocity signal values ​​corresponding to the number of input layers are input into the RBF radial basis neural network prediction model to obtain the velocity signal values ​​of the floor plate of the seat suspension at future moments.

[0087] In one possible embodiment, step 3 specifically involves:

[0088] Step 31. The speed signal of the seat suspension floor plate predicted by the RBF neural network prediction model is transmitted to the Fuzzy-PID control module.

[0089] Step 32. Set the desired seat suspension speed signal for the Fuzzy-PID control module and pass this signal to the Fuzzy-PID control module.

[0090] Step 33. The Fuzzy-PID control module calculates the deviation using the signals obtained in steps 31 and 32, and outputs the control force of the air spring using the Fuzzy-PID control algorithm.

[0091] like Figure 4 As shown, in one possible embodiment, step 33 specifically includes the following steps:

[0092] Step 331. Compare the predicted speed signal of the seat suspension upper plate with the desired seat suspension speed signal, calculate the error signal e and its error change rate ec, and use them as input variables for the Fuzzy-PID control module. The output fuzzy control variable is ΔK. p ΔK i ΔKd .

[0093] Step 332. The error signal e and its error change rate ec are used as input variables of the Fuzzy-PID control module. After the quantization factor is applied, the corresponding fuzzy variables E and EC are obtained, and the basic physical domain is transformed into a fuzzy domain. Specifically, the fuzzy subsets are taken as {NB, NM, NS, O, PS, PM, PB}, and the fuzzy terms are divided into 7 levels, namely -3, -2, -1, 0, 1, 2, 3. The membership function of each fuzzy subset is selected as a Gaussian membership function to describe it. Among them, NB represents negative large, NM represents negative medium, NS represents negative small, PS represents positive small, PM represents positive medium, and PB represents positive large.

[0094] Step 333. Perform fuzzy inference using fuzzy rules, that is, look up the fuzzy inference table based on the membership degrees of the two input signals, the error signal e and its rate of change of error ec, to obtain the output fuzzy control vector ΔK. p ΔK i ΔK d Fuzzy rules are shown in Tables 1 to 3.

[0095] Table 1 shows ΔK p Fuzzy control rule table:

[0096]

[0097] Table 2 shows ΔK i Fuzzy control rule table:

[0098]

[0099]

[0100] Table 3 shows ΔK d Fuzzy control rule table:

[0101]

[0102] Step 334. Apply the centroid method to the output fuzzy control vector ΔK. p ΔK i ΔK d After defuzzification processing, the parameters of the traditional PID controller are adjusted to achieve adaptive adjustment of the proportional, integral, and derivative coefficients of the PID controller, enabling the control system to achieve precise control and improve the vibration isolation performance of the air spring adjustable seat suspension and the ride comfort of the driver and passengers.

[0103] This method takes into account the potential deterioration of control performance due to response delay and overcharging / over-discharging, as well as the possibility of excessively high or low gas pressure. It utilizes an RBF radial basis function neural network prediction model to process the collected data and predict the speed of the floor plate on the seat suspension. The predicted speed signal is then used as the input to the Fuzzy-PID control module to predict the corresponding control force of the air spring. This allows the solenoid valve controlling the inflation or deflation to be opened in advance, thus inflating or deflating the air spring ahead of time, thereby changing the air spring force and minimizing the impact of the inflation / deflation time lag.

[0104] Overcoming the shortcomings of traditional PID controllers, a rule base for fuzzy controllers (i.e., Tables 1 to 3) is designed. Based on different speed signals, the proportional, integral, and derivative coefficients of the PID controller are adaptively adjusted, enabling the control system to achieve precise control and improve the vibration isolation performance of the air spring adjustable seat suspension and the ride comfort of the driver and passengers.

[0105] like Figure 6 As shown in the embodiments of this application, an air spring adjustable seat suspension control system includes a controller and a memory. The memory stores a computer-readable program, which, when invoked by the controller, can execute the steps of the air spring adjustable seat suspension control method as described in the embodiments of this application.

[0106] In this embodiment of the application, a seat employs an air spring adjustable seat suspension control system as described in this embodiment of the application.

[0107] In this application embodiment, a medium stores a computer-readable program, which, when invoked, can execute the steps of the air spring adjustable seat suspension control method as described in this application embodiment.

[0108] In embodiments of this application, the medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. The medium may be a machine-readable signal medium or a machine-readable storage medium. The medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of the medium include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0109] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for controlling an adjustable air spring seat suspension, characterized in that, Includes the following steps: Step 1. Obtain the time-series velocity signal of the upper plate of the seat suspension in the adjustable air spring seat suspension and the elastic force of the air spring in the adjustable air spring seat suspension, respectively; Step 2. Use the RBF radial basis function neural network prediction model to process the collected time series velocity signals and predict the velocity of the upper plate of the seat suspension; Step 3. Use the predicted speed of the seat suspension upper plate as input to the Fuzzy-PID control algorithm to obtain the corresponding control force of the air spring; specifically including: Step 31. Transmit the speed signal of the seat suspension floor plate predicted by the RBF neural network prediction model to the Fuzzy-PID control module; Step 32. Set the desired seat suspension speed signal for the Fuzzy-PID control module and transmit this signal to the Fuzzy-PID control module; Step 33. The Fuzzy-PID control module calculates the deviation using the signals obtained in steps 31 and 32, and outputs the control force of the air spring using the Fuzzy-PID control algorithm; specifically including: Step 33 specifically includes the following steps: Step 331. Compare the predicted speed signal of the seat suspension upper plate with the desired seat suspension speed signal, calculate the error signal e and its error change rate ec, and use them as input variables for the Fuzzy-PID control module. The output fuzzy control variable is... , , ; Step 332. The error signal e and its error change rate ec are used as input variables of the Fuzzy-PID control module. After the quantization factor is applied, the corresponding fuzzy variables E and EC are obtained, and the basic physical domain is transformed into a fuzzy domain. Specifically, the fuzzy subsets are taken as {NB, NM, NS, O, PS, PM, PB}, and the fuzzy terms are divided into 7 levels, namely -3, -2, -1, 0, 1, 2, 3. The membership function of each fuzzy subset is selected as a Gaussian membership function to describe it. Among them, NB represents negative large, NM represents negative medium, NS represents negative small, PS represents positive small, PM represents positive medium, and PB represents positive large. Step 333. Perform fuzzy inference using fuzzy rules, that is, look up the fuzzy inference table based on the membership degrees of the two input signals, the error signal e and its rate of change of error ec, to obtain the output fuzzy control vector. , , ; Step 334. Apply the centroid method to the output fuzzy control vector. , , After defuzzing, the parameters of the traditional PID controller are adjusted to achieve adaptive adjustment of the PID's proportional, integral, and derivative coefficients. Step 4. When the predicted control force of the air spring is greater than or equal to the spring force of the air spring, open the air inflation solenoid valve to inflate the air spring. When the predicted control force of the air spring is less than the spring force of the air spring, open the air release solenoid valve to change the spring force of the air spring.

2. The air spring adjustable seat suspension control method according to claim 1, characterized in that: The method for constructing the RBF radial basis function neural network prediction model in step 2 is as follows: Set the number of nodes in the input, hidden, and output layers of the RBF neural network; set the time series delay step size of the RBF neural network to d and the prediction step size to p; The activation function of the hidden layer is set to a Gaussian function, as shown in the following formula: ； in: This represents the input corresponding to the i-th hidden layer node in the input sample; This represents the center of the basis function of the i-th hidden layer node; This represents the expansion constant of the i-th hidden layer node; After the data passes through the activation function in the hidden layer, the hidden layer output is as follows: ； in: This represents the output of the i-th hidden layer node; A linear mapping is achieved from the hidden layer to the output layer, as shown in the following formula: ； in: Indicates the output of the output layer; Indicates the number of hidden layer nodes; This represents the weight of the i-th hidden layer node; The parameters of the RBF neural network prediction model are corrected based on the training set data; and when the parameters of the RBF neural network prediction model meet the usage requirements, the optimal parameter set of the RBF neural network prediction model is determined, thus obtaining the trained RBF neural network prediction model.

3. The air spring adjustable seat suspension control method according to claim 2, characterized in that: The number of input layers is 7; the number of nodes in the hidden layers is 7; and the number of nodes in the output layers is 1.

4. The air spring adjustable seat suspension control method according to claim 2, characterized in that: When the prediction error is within the set threshold and has not reached the preset number of iterations, it indicates that the parameters of the RBF neural network prediction model meet the requirements; otherwise, it indicates that the parameters of the RBF neural network prediction model do not meet the requirements, and the parameters of the RBF neural network prediction model are corrected using the gradient descent method; the specific formula is as follows: ； ； ； in, Indicates the learning rate; Indicates the number of iterations; This represents the error function.

5. An air spring adjustable seat suspension control system, characterized in that: It includes a controller and a memory, wherein the memory stores a computer-readable program that, when invoked by the controller, can perform the steps of the air spring adjustable seat suspension control method as described in any one of claims 1 to 4.

6. A type of seat, characterized in that: The air spring adjustable seat suspension control system as described in claim 5 is adopted.

7. A medium, characterized in that: It contains a computer-readable program that, when invoked, performs the steps of the air-spring adjustable seat suspension control method as described in any one of claims 1 to 4.

Images