Mountain leaf transfer vehicle anti-rollover early warning method and system based on data driving

Abstract

The invention provides a mountain leaf transfer vehicle anti-rollover early warning method and system based on data driving, and relates to the technical field of vehicle anti-rollover risk early warning, and the method comprises the following steps: S1, carrying out the joint simulation to generate a neural network training data set; s2, constructing RBF neural network training as a vehicle sprung load parameter estimation model; s3, optimizing parameters of the RBF neural network by using an improved grey wolf optimization algorithm; s4, optimizing the RBF neural network, and performing sprung load prediction based on the optimized RBF neural network; s5, state information in the vehicle driving process is obtained through a vehicle-mounted sensor, the sprung load distribution condition of the vehicle is obtained, and an early warning signal is generated; and S6, building an ROS-based wind power blade mountain transfer vehicle anti-rollover visual platform, and carrying out anti-rollover early warning. According to the method, real-time vehicle rollover prevention early warning is carried out according to the parameters in the vehicle driving process, and the safety in the vehicle transportation process is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle rollover prevention risk warning technology, and in particular to a data-driven mountain blade transport vehicle rollover prevention warning method and system. Background Art

[0002] With the growing demand for energy, the development of new green energy has shown a significant growth trend. Among them, wind power generation is an important source of new green energy, and the layout of wind turbines has gradually extended from coastal areas to inland mountainous areas. The transportation of wind turbines is an important part of the installation of wind turbines, and the consideration of safety during transportation is very critical. Due to its large size and mass, and the winding roads in mountainous areas, wind turbines may face the risk of rollover at any time during transportation, causing irreparable economic losses. Therefore, rollover warning for mountain wind turbine blade transportation vehicles is an important issue that needs to be solved urgently to ensure the safety of goods and operators during transportation.

[0003] The invention with application number 201811296526.3 discloses a vehicle rollover prevention control method, which obtains the air spring load value by obtaining the sensor value change of the electronically controlled air suspension to achieve rollover warning. The invention patent with application number 202311207832.6 discloses a vehicle rollover warning method based on adaptive unscented Kalman filtering, which uses unscented Kalman filtering to estimate the distance from the center of mass to the roll axis, and then judges the roll degree of the vehicle according to the change of the lateral load, thereby achieving rollover warning. The invention with application number 202310668601.9 discloses a heavy-duty truck curve rollover warning method based on vehicle-road collaboration, which achieves rollover warning effect by calculating the roll degree of the heavy-loaded vehicle and the maximum safe speed of the corner.

[0004] At present, there are few studies on the rollover warning of oversized transport vehicles. The rollover warning of vehicles is mainly focused on conventional vehicles. Some studies have also been conducted on the rollover warning of heavy-loaded vehicles, but most of them remain in the theoretical stage without many practical solutions. At the same time, the current methods for vehicle rollover warning are mainly based on physical model methods, which have high requirements for the accuracy of model establishment and accurate identification of vehicle parameters. Summary of the invention

[0005] In order to solve the deficiencies of the above-mentioned prior art, the purpose of the present invention is to provide a data-driven anti-rollover warning method and system for mountain blade transport vehicles, which uses sensors arranged on the vehicle to collect parameters during the vehicle's driving process, such as the vehicle's longitudinal speed, front wheel angle, yaw angular velocity, lateral acceleration, roll angle, roll angular velocity, etc. Based on the improved and optimized RBF neural network training, the vehicle's state information during driving is obtained through the on-board sensors to obtain the vehicle's sprung load distribution, and the rollover time TTR is obtained, and a rollover warning is generated in combination with the rollover prefabrication. The anti-rollover warning method has the advantages of being real-time, accurate, and easy to implement, and can more accurately assess the rollover risk of the vehicle.

[0006] Specifically, the present invention provides a data-driven anti-rollover warning method for a mountain blade transport vehicle, which is characterized in that it includes the following steps:

[0007] S1, joint simulation generates data set;

[0008] S2. Construct RBF neural network training as the vehicle sprung load parameter estimation model to determine the vehicle speed v x , the front wheel steering angle δ1, the trailer rear wheel active steering angle δ2 and the upper load working mechanism rotation angle θ are used as the input of the prediction model; the sprung load F of each wheel of the vehicle is determined i is the output of the prediction model;

[0009] S3, using the improved grey wolf optimization algorithm to optimize the parameters of the RBF neural network, and defining the levels as α, β, γ and η when solving the problem, with the levels decreasing in sequence, where α represents the optimal solution, β, γ, and η represent the second solution, the third solution, and the fourth solution respectively;

[0010] In the iterative solution process, an adaptive convergence factor a is introduced to solve the problem of linear decrease of the convergence factor during the iteration process:

[0011]

[0012] Among them, rand represents a random number between 0 and 1; t represents the current number of iterations, and T represents the maximum number of iterations;

[0013] The Gauchy distribution function is introduced to optimize the optimal solution X(t+1). The optimized expression is:

[0014]

[0015] Among them, X1, X2, and X3 are the position vectors of the optimal solution α, the second solution β, and the third solution γ respectively;

[0016] S4, optimizing the RBF neural network and predicting the sprung load based on the optimized RBF neural network, specifically including the following sub-steps:

[0017] S41. Determine the number of neurons in the hidden layer:

[0018]

[0019] Among them, τ1, τ2 represent the number of neurons in the input layer, q∈[1,20];

[0020] Represents the number of hidden neurons;

[0021] S42, data preprocessing, dividing the data into training samples and test samples and performing standardization processing;

[0022] S43, setting parameters of the gray wolf optimization algorithm, including the maximum number of iterations and the number of gray wolves;

[0023] S44, calculating the fitness value of each gray wolf, and sorting the gray wolves according to the fitness value;

[0024] Step S45: updating the parameters based on step S3 until the iteration is completed, and optimizing the parameters of the network according to the above parameters;

[0025] S46, predicting the sprung load based on the optimized RBF neural network;

[0026] S5. Obtain the state information of the vehicle during driving through the vehicle-mounted sensor to obtain the distribution of the vehicle's sprung load, obtain the rollover time TTR, and generate a rollover warning signal based on the rollover time TTR and the rollover threshold. The rollover threshold is calculated as follows:

[0027]

[0028] Among them, k i represents the weight coefficient θ represents the rotation angle of the upper loading device, φ1, represents the yaw angle of the transport chassis tractor and trailer;

[0029] S6. Build a ROS-based wind turbine blade mountain transport vehicle anti-rollover visualization platform to provide anti-rollover warning.

[0030] Preferably, in step S1, S1 utilizes Recurdyn multi-body dynamics simulation software and TruckSim vehicle dynamics simulation software to jointly simulate and generate a neural network training data set.

[0031] Preferably, step S1 specifically includes the following sub-steps:

[0032] S11. Build a three-dimensional model of the loading mechanism of the wind turbine blade mountain transport vehicle in Recurdyn multi-body dynamics simulation software according to the actual vehicle parameters;

[0033] S12. Build a vehicle dynamics model of a wind turbine blade mountain transporter chassis semi-trailer tractor in TruckSim software based on real vehicle parameters;

[0034] S13. Build Recurdyn & TruckSim & in MATLAB / Simulink software

[0035] MATLAB / Simulink joint simulation environment;

[0036] S14, sending the rotation angle information of the upper load working mechanism to Recurdyn through Simulink, and sending the vehicle control parameters to TruckSim through Simulink;

[0037] S15, Recurdyn performs actions according to the received rotation angle information of the operating mechanism, generates a corresponding rotation angle, and calls the output interface to send the load parameters of the calibration point to TrcukSim; TruckSim outputs the corresponding vehicle spring load value according to the received information;

[0038] S16, generating a data set for estimating vehicle sprung load parameters according to the data of steps S11-S15;

[0039] S17, screening and filtering the data set generated in step S16 to finally generate a data set.

[0040] Preferably, step S2 specifically includes the following steps:

[0041] S21. Establish a nonlinear network architecture that describes the sprung load change as a vehicle sprung load parameter estimation model. The nonlinear network architecture adopts a three-layer structure, which includes an input layer, a hidden layer, and an output layer. The input vector is nonlinearly mapped to the hidden space through the activation function of the hidden layer. The output of the prediction model is the linear weighted sum of the outputs of the hidden layer neurons.

[0042] S22, determining the input and output of the vehicle sprung load parameter estimation model;

[0043] S23, select the Gaussian function as the activation function, use the approximate output of the RBF neural network to obtain the estimated sprung load value, and further calculate the actual sprung load value:

[0044]

[0045] in, represents the estimated sprung load value, c i and σ i represents the center and width of the neuron, ω represents the weight between the output and the hidden layer, Indicates the actual sprung load value.

[0046] Preferably, step S3 specifically includes the following sub-steps:

[0047] S31. Define boundary prey. During hunting, the behavior of wolf hunters is defined as follows:

[0048] D=|C·X P (t)-X(t)|;

[0049] The gray wolf's position update formula is:

[0050] X(t+1)=X P (t)-A·D

[0051] Where D represents the distance between the individual and the prey, t represents the current number of iterations, A and C are coefficient vectors, and X P and X are the position vectors of the prey and the gray wolf respectively, and the calculation formulas of A and C are as follows:

[0052] A=2a·r1-a

[0053] C=2·r2

[0054]

[0055] Where a is the convergence factor. As the number of iterations increases, the convergence factor decreases linearly from 2 to 0. r1 and r2 are random numbers between [0,1].

[0056] S32. When the gray wolf determines the location of the prey, β and γ surround the prey under the guidance of α. The mathematical model of the gray wolf tracking the prey location is described as follows:

[0057] D α =|C1·X α -X|

[0058] D β =|C2·X β -X|

[0059] D γ =|C3·X γ -X|

[0060] X1=X α -A1·D α

[0061] X2=X β -A2·Dβ

[0062] X3=X γ -A3·D γ

[0063]

[0064] Among them, D α , D β , D γ Respectively represent the distances between α, β, γ and other individuals; X α , X β and X γ represents the current position vector of α, β, γ; C1, C2, C3 are random vectors; X is the current position of the gray wolf;

[0065] S33. During the iteration process, when the a value decreases linearly from 2 to 0, the corresponding a changes from -a to a. When |a|<1, the wolf attacks the prey; when |A|>1, the gray wolf is separated from the prey.

[0066] Preferably, the standardized expression in step S42 is:

[0067]

[0068] Among them, X k represents the Kth eigenvector;

[0069] In step S44, the fitness function is expressed as follows:

[0070]

[0071] Among them, y(i), Represent the actual value and predicted value respectively; i represents the i-th sample, and n represents the total number of samples.

[0072] Preferably, step S5 specifically includes the following sub-steps:

[0073] Step S51, selecting the vehicle rollover index, and defining the vehicle's lateral load transfer rate LTR as:

[0074]

[0075] Step S52: using the trained vehicle sprung load parameter estimation model, returning the vehicle rollover time TTR according to the vehicle state information;

[0076] Step S53: Generate a rollover warning signal based on the rollover time TTR and the rollover threshold, set the alarm rollover time T=T1*ε, T1 is set according to demand, ε is the rollover threshold, and when the rollover time TTR is less than or equal to T, output the rollover warning risk.

[0077] Preferably, step S6 specifically includes the following sub-steps:

[0078] S61. Use 3D modeling software to establish a three-dimensional model of a wind turbine blade mountain transport vehicle;

[0079] S62. Constrain the three-dimensional model in the ROS space and establish a three-dimensional model of a wind turbine blade mountain transporter with multiple degrees of freedom;

[0080] S63, collecting signals through vehicle-mounted sensors, and inputting the collected signals into the vehicle-mounted visualization embedded system;

[0081] S64, reading the corresponding signal through the ROS node, inputting the signal into the interface for controlling the movement of the three-dimensional model, synchronizing the position and posture information of the vehicle in real time in the ROS space, and displaying it on the screen;

[0082] S65, displaying the vehicle status information and the parameter change information of the sprung load on the screen;

[0083] S66. According to the set threshold, the indicator light of the embedded system will display different colors, green for safety, yellow for potential rollover risk, and red for imminent rollover;

[0084] S67. When there is a risk of rollover, the lights of the vehicle-mounted embedded system flash rapidly and the buzzer sounds an alarm, achieving a rollover warning effect.

[0085] Preferably, the vehicle control parameters in step S14 include vehicle speed and steering wheel angle.

[0086] On the other hand, the present invention provides an early warning system for the above-mentioned data-driven anti-rollover early warning method for mountain blade transport vehicles, which includes a data set generation unit, a vehicle sprung load parameter estimation model generation unit, an RBF neural network parameter optimization unit, a sprung load prediction unit, a rollover warning signal generation unit, and a rollover warning unit;

[0087] The data set generation unit is used for joint simulation to generate data sets; the vehicle sprung load parameter estimation model generation unit is used for constructing RBF neural network training as a vehicle sprung load parameter estimation model; the RBF neural network parameter optimization unit is used for optimizing the parameters of the RBF neural network using the improved grey wolf optimization algorithm; the sprung load prediction unit is used for optimizing the RBF neural network and predicting the sprung load based on the optimized RBF neural network; the rollover warning signal generation unit is used for obtaining the status information of the vehicle during driving through the on-board sensor to obtain the sprung load distribution of the vehicle and generate a warning signal; the rollover warning unit is used to build a ROS-based wind turbine blade mountain transport vehicle anti-rollover visualization platform for anti-rollover warning.

[0088] Compared with the prior art, the present invention has the following beneficial effects:

[0089] (1) The method of the present invention uses sensors arranged on the vehicle to collect parameters during the vehicle's driving process, such as the vehicle's longitudinal speed, front wheel angle, yaw angular velocity, lateral acceleration, roll angle, roll angular velocity, etc. Based on the improved and optimized RBF neural network training, the vehicle's state information during the vehicle's driving process is obtained through the vehicle-mounted sensors to obtain the vehicle's sprung load distribution, and the rollover time TTR is obtained. Combined with the rollover prefabrication, a rollover warning is generated. The anti-rollover warning method has the advantages of being real-time, accurate, and easy to implement, and can more accurately assess the rollover risk of the vehicle.

[0090] (2) The method of the present invention utilizes Recurdyn multi-body dynamics simulation software and TruckSim vehicle dynamics simulation software to jointly simulate and generate a neural network training data set, which can ensure the authenticity and diversity of the training data and further improve the accuracy of the prediction.

[0091] (3) The method of the present invention uses the improved grey wolf optimization algorithm (IGWO) to optimize the parameters of the RBF neural network to improve the estimation accuracy of the model. The optimized RBF neural network is used to predict vehicle load, which can greatly improve the prediction accuracy and ensure the accuracy of rollover warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0092] Figure 1 is a flow chart of the method of the present invention;

[0093] Figure 2 It is a schematic diagram of the overall process of the method of the present invention;

[0094] Figure 3 It is a schematic diagram of the data generation method of the present invention;

[0095] Figure 4 It is a schematic diagram of the neural network structure of the present invention;

[0096] Figure 5 It is a schematic diagram of the structure of the vehicle-mounted embedded system of the present invention. DETAILED DESCRIPTION

[0097] Hereinafter, embodiments of the present invention will be described with reference to the drawings.

[0098] like Figure 2 As shown, the present invention provides a data-driven anti-rollover warning method for mountain blade transport vehicles. It uses Recurdyn multi-body dynamics simulation software and TruckSim vehicle dynamics simulation software to jointly simulate and generate a neural network training data set; generates a vehicle sprung load parameter estimation model through improved RBF neural network training; deploys the parameter estimation model on the vehicle, obtains the vehicle's sprung load distribution through the vehicle's onboard sensor information to obtain the vehicle's state information during driving, and calculates the rollover time TTR, and generates the vehicle's posture and load distribution information in the visual embedded system of the cockpit according to the sensor parameters, thereby achieving an anti-rollover warning effect.

[0099] Specifically, Figure 1 As shown, the steps of the data-driven mountain blade transport vehicle anti-rollover warning method of the present invention are as follows:

[0100] S1, dataset generation, e.g. Figure 3 As shown, the specific steps include:

[0101] S11. Build a 3D model of the loading mechanism of the wind turbine blade mountain transport vehicle in the Recurdyn multi-body dynamics simulation software based on the actual vehicle parameters.

[0102] S12. Build the vehicle dynamics model of the wind turbine blade mountain transporter chassis semi-trailer tractor in TruckSim software based on the actual vehicle parameters.

[0103] S13. Build Recurdyn & TruckSim & in MATLAB / Simulink software

[0104] MATLAB / Simulink joint simulation environment.

[0105] S14. Send the rotation angle information of the upper load working mechanism to Recurdyn through Simulink; send the vehicle speed, steering wheel angle and other control parameters to TruckSim.

[0106] S15. Recurdyn performs actions according to the received rotation angle information of the operating mechanism, generates a corresponding rotation angle, and calls the output interface to send the load parameters of the calibration point to TruckSim; TruckSim outputs the corresponding vehicle spring load value according to the received information.

[0107] S16. Generate a data set for estimating vehicle sprung load parameters based on the data from steps S11 to S15.

[0108] S17, screening and filtering the data set generated in step S16, and finally generating a training set that can be used for neural network training.

[0109] S2. Construct the RBF neural network structure. The RBF neural network structure is as follows: Figure 4 As shown, the specific steps include:

[0110] S21. Establish a nonlinear network architecture to describe the change of sprung load. A three-layer structure is used, including input layer, hidden layer, and output layer. The input vector is nonlinearly mapped to the hidden space through the activation function of the hidden layer. The output of the prediction model is the linear weighted sum of the outputs of the neurons in the hidden layer.

[0111] S22, select four variables as the input of the network: vehicle speed v x , front wheel steering angle δ1, trailer rear wheel active steering angle δ2, upper load mechanism rotation angle θ; the output of the network is selected as the sprung load F of each wheel of the vehicle i .

[0112] S23. Select the Gaussian function as the activation function, and the approximate output of the RBF neural network is:

[0113]

[0114] in, represents the output of the hidden neuron, c i and σ i represents the center and width of the neuron, and ω represents the weight between the output and the hidden layer.

[0115] S3. Use the improved grey wolf optimization algorithm (IGWO) to optimize the parameters of the RBF neural network to improve the estimation accuracy of the model. There is a strict hierarchical distribution relationship between wolves. The social levels are divided into α, β, γ and η, and the levels decrease in sequence. Among them, α is the alpha wolf, representing the solution with the highest fitness value, and β, γ, and η represent the second, third, and other solutions respectively. It includes the following sub-steps:

[0116] S31. Define boundary prey:

[0117] During the hunting process, the actions of wolf hunters are defined as follows:

[0118] D=|C·X P (t)-X(t)|.

[0119] The gray wolf's position update formula is:

[0120] X(t+1)=X P (t)-A·D

[0121] Where D represents the distance between the individual and the prey, t represents the current number of iterations, A and C are coefficient vectors, and X P and X are the position vectors of the prey and the wolf respectively. The calculation formulas of A and C are as follows:

[0122] A=2a·r1-a

[0123] C=2·r2

[0124]

[0125] Where a is the convergence factor. As the number of iterations increases, the convergence factor decreases linearly from 2 to 0. r1 and r2 are random numbers between [0,1].

[0126] S32. Hunting:

[0127] Wolves are able to identify prey and surround them. When the gray wolf determines the location of the prey, β and γ surround the prey under the guidance of α. The mathematical model of the gray wolf tracking the prey location is described as follows:

[0128] D α =|C1·X α -X|

[0129] D β =|C2·X β -X|

[0130] D γ =|C3·X γ -X|

[0131] X1=X α -A1·D α

[0132] X2=X β -A2·D β

[0133] X3=X γ -A3·D γ

[0134]

[0135] Among them, D α, D β , D γ Respectively represent the distances between α, β, γ and other individuals; X α , X β and X γ represents the current position vector of α, β, γ; C1, C2, C3 are random vectors; X is the current position of the gray wolf

[0136] Introducing an adaptive convergence factor a to solve the problem of linear decrease of the convergence factor during the iteration process

[0137]

[0138] Among them, rand represents a random number between 0 and 1.

[0139] The Gauchy distribution function is introduced to optimize the optimal solution X(t+1). The optimized expression is:

[0140]

[0141] S33. Attacking prey:

[0142] When the prey stops moving, the gray wolf completes the hunting process by attacking. In order to simulate the process of approaching the prey, the decline of a is gradually controlled. At this time, the fluctuation amplitude of A is reduced accordingly. That is to say, during the iteration process, when the value of a decreases linearly from 2 to 0, the corresponding a also changes from -a to a. When |a|<1, the wolf attacks the prey. When |A|>1, the gray wolf is separated from the prey, the global search ability of the algorithm is enhanced, and the most suitable prey is found.

[0143] S4. Estimation of sprung load based on improved RBF neural network

[0144] S41. Determine the number of neurons in the hidden layer

[0145]

[0146] Among them, τ1τ2 represents the number of neurons in the input layer, q∈[1,20].

[0147] S42, data preprocessing, divide the data into training samples and test samples, the standardized expression is:

[0148]

[0149] Among them, X k represents the Kth eigenvector.

[0150] S43. Setting parameters of the gray wolf optimization algorithm, including the maximum number of iterations and the number of gray wolves.

[0151] S44, calculate the fitness value of each gray wolf, and sort the gray wolves according to the fitness value. The fitness function is expressed as follows:

[0152]

[0153] in, represent the actual value and the predicted value respectively.

[0154] S45. Update the parameters according to step S3 until the iteration is completed, and optimize the parameters of the network according to the above parameters.

[0155] S5. Calculation of the rollover warning time of the wind turbine blade transport vehicle based on the change of the spring load:

[0156] S51, selecting a vehicle rollover index, and defining the vehicle's lateral load transfer rate LTR as:

[0157]

[0158] S52: using the trained vehicle sprung load parameter estimation model, returning the vehicle rollover time TTR according to the vehicle status information.

[0159] S53, generating a rollover warning signal based on the rollover time TTR and the rollover threshold, setting the alarm rollover time T=T1*ε, T1 is set according to demand, ε is the rollover threshold, and when the rollover time TTR is less than or equal to T, outputting the rollover warning risk.

[0160] S6. Establish a ROS-based wind turbine blade mountain transport vehicle rollover prevention visualization platform. The visualization platform is as follows: Figure 5 As shown, the specific steps include:

[0161] S61. Use 3D modeling software to build a three-dimensional model of the wind turbine blade mountain transport vehicle.

[0162] S62. Constrain the three-dimensional model in the ROS space to realize a movable multi-degree-of-freedom three-dimensional model of a wind turbine blade mountain transporter.

[0163] S63. Collect signals through vehicle-mounted sensors and input them into the vehicle-mounted visualization embedded system.

[0164] S64. Read the corresponding signal through the ROS node, input it into the interface for controlling the movement of the three-dimensional model, synchronize the vehicle's position information in real time in the ROS space, and display it on the screen.

[0165] S65: Display the vehicle status information and the parameter change information of the sprung load on the screen.

[0166] S66. According to the set threshold, the indicator light of the embedded system will display different colors, green for safety, yellow for potential rollover risk, and red for imminent rollover.

[0167] S67. When there is a risk of rollover, the lights of the vehicle-mounted embedded system flash rapidly and the buzzer sounds an alarm, achieving a rollover warning effect.

[0168] On the other hand, the present invention provides a warning system for the above-mentioned data-driven mountain blade transfer vehicle anti-rollover warning method, which includes a data set generation unit, a vehicle sprung load parameter estimation model generation unit, an RBF neural network parameter optimization unit, a sprung load prediction unit, a rollover warning signal generation unit and a rollover warning unit.

[0169] The data set generation unit is used for joint simulation to generate data sets; the vehicle sprung load parameter estimation model generation unit is used to construct RBF neural network training as the vehicle sprung load parameter estimation model; the RBF neural network parameter optimization unit is used to optimize the parameters of the RBF neural network using the improved Grey Wolf optimization algorithm; the sprung load prediction unit is used to optimize the RBF neural network and predict the sprung load based on the optimized RBF neural network; the rollover warning signal generation unit is used to obtain the vehicle's state information during driving through the on-board sensors to obtain the vehicle's sprung load distribution and generate a warning signal; the rollover warning unit is used to build a ROS-based wind turbine blade mountain transport vehicle anti-rollover visualization platform for anti-rollover warning.

[0170] Specific application examples:

[0171] First, according to the steps shown in the invention, the required sensors are installed on the transport vehicle, and the corresponding data can be read through the display screen of the embedded system to determine whether the sensor installation is normal. When the vehicle is running, the sensor reads the vehicle status information normally. When the vehicle encounters a possible rollover scenario, such as driving on a rugged mountain road, in order to avoid possible collision obstacles, the vehicle's upper operating mechanism rotates to avoid, generating a rotation angle. In the cockpit, the driver monitors the body posture in real time on the screen of the embedded system, and finds that the model in the picture also produces corresponding animation effects with the actual movement of the vehicle. The vehicle continues to drive, and when it reaches a certain section of the road, the body tilts to a certain extent. At this time, the driver shows that the body load distribution is abnormal, and the left wheel load is significantly greater than the right side according to the data sent back by the screen. At the same time, according to the signal detected by the sensor, the indicator light of the vehicle embedded system shows yellow, indicating that there is a risk of rollover, and the actual TTR of the rollover is 30 seconds. The driver receives the corresponding rollover signal and immediately takes action to adjust the vehicle posture. After the execution is completed, the signal light returns to green, the load value returns to normal, and the rollover risk is eliminated.

[0172] The embodiments described above are only descriptions of the preferred implementation modes of the present invention, and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A data-driven anti-rollover warning method for mountain blade transport vehicles, characterized in that: It includes the following steps: S1, joint simulation generates data set; S2. Construct RBF neural network training as the vehicle sprung load parameter estimation model to determine the vehicle speed v x , the front wheel steering angle δ1, the trailer rear wheel active steering angle δ2 and the upper load working mechanism rotation angle θ are used as the input of the prediction model; the sprung load F of each wheel of the vehicle is determined i is the output of the prediction model; S3, using the improved grey wolf optimization algorithm to optimize the parameters of the RBF neural network, and defining the levels as α, β, γ and η when solving the problem, with the levels decreasing in sequence, where α represents the optimal solution, β, γ, and η represent the second solution, the third solution, and the fourth solution respectively; In the iterative solution process, an adaptive convergence factor a is introduced to solve the problem of linear decrease of the convergence factor during the iteration process: Among them, rand represents a random number between 0 and 1; t represents the current number of iterations, and T represents the maximum number of iterations; The Gauchy distribution function is introduced to optimize the optimal solution X(t+1). The optimized expression is: Among them, X1, X2, and X3 are the position vectors of the optimal solution α, the second solution β, and the third solution γ respectively; S4, optimizing the RBF neural network and predicting the sprung load based on the optimized RBF neural network, specifically including the following sub-steps: S41. Determine the number of neurons in the hidden layer: Among them, τ1, τ2 represent the number of neurons in the input layer, q∈[1,20]; N k Represents the number of hidden neurons; S42, data preprocessing, dividing the data into training samples and test samples and performing standardization processing; S43, setting parameters of the gray wolf optimization algorithm, including the maximum number of iterations and the number of gray wolves; S44, calculating the fitness value of each gray wolf, and sorting the gray wolves according to the fitness value; S45, updating the parameters based on step S3 until the iteration is completed, and optimizing the parameters of the network according to the above parameters; S46, predicting the sprung load based on the optimized RBF neural network; S5. Obtain the state information of the vehicle during driving through the vehicle-mounted sensor to obtain the distribution of the vehicle's sprung load, obtain the rollover time TTR, and generate a rollover warning signal based on the rollover time TTR and the rollover threshold. The rollover threshold is calculated as follows: Among them, k i represents the weight coefficient θ represents the rotation angle of the upper loading device, φ1, represents the yaw angle of the transport chassis tractor and trailer; S6. Build a ROS-based wind turbine blade mountain transport vehicle anti-rollover visualization platform to provide anti-rollover warning.

2. The data-driven anti-rollover warning method for mountain blade transport vehicles according to claim 1 is characterized in that: In step S1, Recurdyn multi-body dynamics simulation software and TruckSim vehicle dynamics simulation software are used to jointly simulate and generate a data set.

3. The data-driven anti-rollover warning method for mountain blade transport vehicles according to claim 2 is characterized in that: Step S1 specifically includes the following sub-steps: S11. Build a three-dimensional model of the loading mechanism of the wind turbine blade mountain transport vehicle in Recurdyn multi-body dynamics simulation software according to the actual vehicle parameters; S12. Build a vehicle dynamics model of a wind turbine blade mountain transporter chassis semi-trailer tractor in TruckSim software based on real vehicle parameters; S13. Build Recurdyn & TruckSim & in MATLAB / Simulink software MATLAB / Simulink joint simulation environment; S14, sending the rotation angle information of the upper load working mechanism to Recurdyn through Simulink, and sending the vehicle control parameters to TruckSim through Simulink; S15, Recurdyn performs actions according to the received rotation angle information of the operating mechanism, generates a corresponding rotation angle, and calls the output interface to send the load parameters of the calibration point to TrcukSim; TruckSim outputs the corresponding vehicle spring load value according to the received information; S16, generating a data set for estimating vehicle sprung load parameters according to the data of steps S11-S15; S17, screening and filtering the data set generated in step S16, and finally generating a training data set for neural network training.

4. The data-driven anti-rollover warning method for mountain blade transport vehicles according to claim 1 is characterized in that: Step S2 specifically includes the following steps: S21. Establish a nonlinear network architecture that describes the sprung load change as a vehicle sprung load parameter estimation model. The nonlinear network architecture adopts a three-layer structure, which includes an input layer, a hidden layer, and an output layer. The input vector is nonlinearly mapped to the hidden space through the activation function of the hidden layer. The output of the prediction model is the linear weighted sum of the outputs of the hidden layer neurons. S22, determining the input and output of the vehicle sprung load parameter estimation model; S23, select the Gaussian function as the activation function, use the approximate output of the RBF neural network to obtain the estimated sprung load value, and further calculate the actual sprung load value: in, represents the estimated sprung load value, c i and σ i represents the center and width of the neuron, ω represents the weight between the output and the hidden layer, Indicates the actual sprung load value.

5. The data-driven anti-rollover warning method for mountain blade transport vehicles according to claim 1 is characterized in that: Step S3 specifically includes the following sub-steps: S31. Define boundary prey. During hunting, the behavior of wolf hunters is defined as follows: D=|C·X P (t)-X(t)|; The gray wolf's position update formula is: X(t+1)=X P (t)-A·D Where D represents the distance between the individual and the prey, t represents the current number of iterations, A and C are coefficient vectors, and X P and X are the position vectors of the prey and the gray wolf respectively, and the calculation formulas of A and C are as follows: A=2a·r1-a C＝2·r2 Where a is the convergence factor. As the number of iterations increases, the convergence factor decreases linearly from 2 to 0. r1 and r2 are random numbers between [0,1]. S32. When the gray wolf determines the location of the prey, β and γ surround the prey under the guidance of α. The mathematical model of the gray wolf tracking the prey location is described as follows: D α =|C1·X α -X| D β =|C2·X β -X| D γ =|C3·X γ -X| X1=X α -A1·D α X2=X β -A2·D β X3=X γ -A3·D γ Among them, D α , D β , D γ Respectively represent the distances between α, β, γ and other individuals; X α , X β and X γ represents the current position vector of α, β, γ; C1, C2, C3 are random vectors; X is the current position of the gray wolf; S33. During the iteration process, when the a value decreases linearly from 2 to 0, the corresponding a changes from -a to a. When |a|<1, the wolf attacks the prey; when |A|>1, the gray wolf is separated from the prey.

6. The data-driven mountain blade transport vehicle rollover prevention warning method according to claim 1 is characterized in that: The standardized expression in step S42 is: Among them, X k represents the Kth eigenvector; In step S44, the fitness function is expressed as follows: Among them, y(i), Represent the actual value and predicted value respectively; i represents the i-th sample, and n represents the total number of samples.

7. The data-driven mountain blade transport vehicle rollover prevention warning method according to claim 1 is characterized in that: Step S5 specifically includes the following sub-steps: Step S51, selecting the vehicle rollover index, and defining the vehicle's lateral load transfer rate LTR as: Step S52: using the trained vehicle sprung load parameter estimation model, returning the vehicle rollover time TTR according to the vehicle state information; Step S53: Generate a rollover warning signal based on the rollover time TTR and the rollover threshold, set the alarm rollover time T=T1*ε, T1 is set according to demand, ε is the rollover threshold, and when the rollover time TTR is less than or equal to T, output the rollover warning risk.

8. The data-driven anti-rollover warning method for mountain blade transport vehicles according to claim 1 is characterized in that: Step S6 specifically includes the following sub-steps: S61. Use 3D modeling software to establish a three-dimensional model of a wind turbine blade mountain transport vehicle; S62. Constrain the three-dimensional model in the ROS space and establish a three-dimensional model of a wind turbine blade mountain transporter with multiple degrees of freedom; S63, collecting signals through vehicle-mounted sensors, and inputting the collected signals into the vehicle-mounted visualization embedded system; S64, reading the corresponding signal through the ROS node, inputting the signal into the interface for controlling the movement of the three-dimensional model, synchronizing the position and posture information of the vehicle in real time in the ROS space, and displaying it on the screen; S65, displaying the vehicle status information and the parameter change information of the sprung load on the screen; S66. According to the set threshold, the indicator light of the embedded system will display different colors, green for safety, yellow for potential rollover risk, and red for imminent rollover; S67. When there is a risk of rollover, the lights of the vehicle-mounted embedded system flash rapidly and the buzzer sounds an alarm, achieving a rollover warning effect.

9. The data-driven anti-rollover warning method for mountain blade transport vehicles according to claim 3 is characterized in that: The vehicle control parameters in step S14 include vehicle speed and steering wheel angle.

10. An early warning system for the data-driven anti-rollover early warning method for mountain blade transport vehicles according to claim 1, characterized in that: It includes a data set generation unit, a vehicle sprung load parameter estimation model generation unit, an RBF neural network parameter optimization unit, a sprung load prediction unit, a rollover warning signal generation unit, and a rollover warning unit; The data set generation unit is used for joint simulation to generate data sets; the vehicle sprung load parameter estimation model generation unit is used for constructing RBF neural network training as a vehicle sprung load parameter estimation model; the RBF neural network parameter optimization unit is used for optimizing the parameters of the RBF neural network using an improved grey wolf optimization algorithm; the sprung load prediction unit is used for optimizing the RBF neural network and predicting the sprung load based on the optimized RBF neural network; The rollover warning signal generating unit is used to obtain the status information of the vehicle during driving through the vehicle-mounted sensor to obtain the vehicle's sprung load distribution and generate a warning signal; the rollover warning unit is used to build a ROS-based wind turbine blade mountain transport vehicle anti-rollover visualization platform to perform anti-rollover warning.

Images