Unmanned walking type excavator body posture stability control method based on neural network

Abstract

The invention provides a neural network-based unmanned walking type excavator body posture stability control method. The method comprises the following steps: scanning the terrain in front of a walking type excavator through a laser radar and a camera to generate terrain point cloud data, and processing the terrain point cloud data to obtain terrain excitation data; the terrain excitation data and the walking leg included angle of the walking excavator serve as input, a neural network attitude prediction model of the walking excavator is constructed, and the pitch angle, the roll angle and the vertical displacement of a walking excavator body are output; based on the vehicle body posture change of the walking excavator, a vehicle body stability optimal control model of the walking excavator is constructed and solved, so that the included angle between a walking leg and a chassis is optimized, and stable control over the vehicle body posture under the complex terrain condition is achieved. The system can efficiently adapt to the complex terrain environment, the stability and safety of the unmanned walking type excavator in autonomous operation are improved, and the system has important significance in improving the operation efficiency and quality.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of engineering machinery control and intelligent optimization, and in particular to a method for stabilizing the body posture of an unmanned walking excavator based on a neural network. Background Art

[0002] As an all-terrain, multi-purpose walking excavator, crawler excavator is widely used for operations in complex terrain environments such as mountains, hills, and forests.

[0003] Compared with traditional crawler excavators, the chassis of walking excavators uses four independent walking mechanisms to adjust the balance of the body, thus greatly improving the applicability and stability of the excavator. The stability of walking excavators is crucial to operational safety and construction efficiency. Since the traditional kinematic model contains nonlinear terms of sine and cosine functions, analytical solutions can be obtained by simplifying the processing under small-angle obstacles, but such simplification cannot be performed at large pitch and roll angles, resulting in a long time to solve the numerical solution in the optimization model, which in turn makes it difficult for walking excavators to achieve real-time adjustment of the body balance under complex terrain conditions. Summary of the invention

[0004] In view of this, an embodiment of the present invention provides a method for stabilizing the body posture of an unmanned walking excavator based on a neural network, aiming to solve the problems of insufficient stability and inaccurate posture control of existing walking excavators in complex terrains.

[0005] According to the first aspect of an embodiment of the present invention, a method for controlling the body posture stability of an unmanned walking excavator based on a neural network is provided, comprising: S1, scanning the terrain in front of the walking excavator by means of a laser radar and a camera to generate terrain point cloud data, processing the terrain point cloud data to obtain terrain excitation data; S2, taking the terrain excitation data and the walking leg angle of the walking excavator as input, constructing a neural network posture prediction model for the walking excavator, and outputting the pitch angle, roll angle and vertical displacement of the walking excavator body to predict the body posture change of the walking excavator; S3, constructing an optimal control model for the body stability of the walking excavator based on the body posture change of the walking excavator, comprising: optimizing variables, objective function and model constraints; S4, assigning values ​​to parameters in the optimal control model for the body stability; S5, assigning values ​​to the variables in the optimal control model for the body stability; Discretization processing is performed at the matching points to obtain processed variables; S6, the processed variables at the matching points are assigned values ​​to obtain the vehicle body posture, hydraulic cylinder angle and hydraulic cylinder speed time series trajectory; S7, according to the vehicle body posture, hydraulic cylinder angle and hydraulic cylinder speed time series trajectory, the objective function and constraint violation degree of the vehicle body stability optimal control model under the current driving trajectory are calculated, including the hydraulic cylinder extension length, hydraulic cylinder extension speed, and the distance from the tire center to the ground; S8, it is determined whether the vehicle body stability optimal control model meets the optimization convergence conditions. If so, the optimal control variables of the vehicle body stability optimal control model are output. If not, go to step S9; S9, the state variables and control variables of the vehicle body stability optimal control model are optimized in a single-step iterative manner, and go to step S6 until the optimal control variables of the vehicle body stability optimal control model are output to achieve stability control of the walking excavator body.

[0006] In one implementation, the terrain excitation data in step S1 includes slope, smoothness and concavity.

[0007] In another implementation, in step S1, the terrain point cloud data is processed by at least one of statistical filtering, radius filtering denoising, Voxel Grid filtering downsampling, and RANSAC algorithm plane segmentation.

[0008] In another implementation, in step S2, the method for establishing the neural network posture prediction model of the walking excavator is as follows: based on the terrain excitation data and the walking leg angle of the excavator, a training data set of the neural network posture prediction model is constructed, the input data of the neural network posture prediction model includes the terrain excitation (w1, w2, w3, w4) and the angle between the walking leg and the chassis (θ1, θ2, θ3, θ4), and the output data is the body posture parameters, including: the pitch angle θ s , roll angle and vertical displacement z sThe neural network posture prediction model adopts a multi-layer feedforward neural network structure, including an input layer, several hidden layers and an output layer. The number of nodes in the input layer is n+4, where n represents the number of terrain excitation features, and the 4 nodes represent the angle between the walking leg and the chassis. The hidden layer is composed of multiple layers of neurons, and the activation function uses ReLU. The output layer contains 3 nodes, corresponding to the pitch angle θ s , roll angle and vertical displacement z s ; Define the loss function and optimization target, train the neural network posture prediction model through supervised learning, select the mean square error loss function to minimize the error between the predicted output of the neural network posture prediction model and the actual posture data. The mean square error loss function is defined as:

[0009]

[0010] in, Represents the output value predicted by the neural network posture prediction model, y i is the true posture value, m is the number of samples; the optimization method adopts the Adam optimization algorithm, and the weights of the neural network posture prediction model are iteratively updated according to the gradient information to accelerate convergence; the neural network posture prediction model is trained for multiple rounds and the weight parameters are updated to reduce the error between the predicted output and the true value; the regularization and early stopping strategies are adopted, and the optimal model structure and hyperparameters of the neural network posture prediction model are determined through cross-validation to prevent the neural network posture prediction model from overfitting; after the training is completed, the neural network posture prediction model can accurately predict the body posture changes of the walking excavator according to the input terrain excitation data and the walking leg angle.

[0011] In another implementation, the Adam optimization algorithm dynamically adjusts the learning rate based on the gradient average of the first-order and second-order moments to accelerate the speed at which the neural network posture prediction model reaches the optimal solution.

[0012] In another implementation, in step S3, the steps for constructing the optimal control model of the body stability of the walking excavator are as follows: taking the walking leg angle as the optimization variable and the stability of the body posture as the objective function, the optimal control model of the body stability is constructed, and the objective function is expressed as:

[0013]

[0014] Among them, θ s represents the pitch angle, is the roll angle, z s represents the vertical displacement, t f Indicates the excavator operation time;

[0015] The model constraints include: the extension and retraction speed v of the hydraulic cylinders of the four supporting legs of the walking excavator during the posture adjustment process hi (i=1,2,3,4) to ensure that the maximum allowable lower limit v hmin and the maximum allowed upper limit v hmax between:

[0016]

[0017] During the movement, the center of the four-leg tires of the walking excavator is displaced by z ui (i=1,2,3,4) and the roadside elevation w corresponding to each wheel i The difference is kept within the maximum permissible lower limit ε min and the maximum permissible upper limit ε max between:

[0018]

[0019] Therefore, the optimal control model of the body stability of the walking excavator is expressed as:

[0020] findθ i (i=1,2,3,4)

[0021]

[0022] sC ij ≤0(i=1,2,3,4,j=1,2,3,4)

[0023] t min ≤t f ≤t max

[0024] Among them, θ i (i=1,2,3,4) represents the walking leg angle, θ s represents the pitch angle, is the roll angle, z s represents the vertical displacement, t f represents the running time, t min represents the minimum running time, t max Indicates the maximum running time.

[0025] In another implementation, in step S5, the steps of discretizing the optimal control model of vehicle body stability are as follows: first, the integral term in the objective function is discretized, and the time interval |0,t f | is divided into N small intervals, and the integral is approximately in the form of summation:

[0026]

[0027] in, For each discrete walking time length, the optimization problem of the vehicle body stability optimal control model is transformed into solving the optimal combination of discrete variables through discretization; then the genetic algorithm is used to solve it, including: the walking leg angle θ i (i=1,2,3,4) is encoded as the gene of the individual, and the population is initialized so that each individual represents an angle combination; by defining the objective function The fitness of each individual is calculated, and excellent individuals are retained and generated in the selection, crossover and mutation operations. The penalty function method is used to reduce the fitness of individuals that do not meet the constraints to meet the walking leg angle range constraints and angular velocity constraints. The vehicle body stability optimal control model is iteratively evolved to obtain the optimal solution for the walking leg angle combination and realize the stability control of the walking excavator body.

[0028] According to the second aspect of an embodiment of the present invention, a body posture stability control system for an unmanned walking excavator based on a neural network is provided, comprising: a data acquisition module, for scanning the terrain in front of the walking excavator through a laser radar and a camera, generating terrain point cloud data, processing the terrain point cloud data, and obtaining terrain excitation data; a neural network posture prediction model module, for taking the terrain excitation data and the walking leg angle of the walking excavator as input, constructing a neural network posture prediction model of the walking excavator, and outputting the pitch angle, roll angle and vertical displacement of the walking excavator body to predict the body posture change of the walking excavator; a body stability optimal control model module, for constructing a body stability optimal control model of the walking excavator based on the body posture change of the walking excavator, comprising: optimizing variables, objective functions, model constraints, and assigning parameters in the body stability optimal control model; body stability optimal control model module; The optimal control model solving module is used to discretize the variables of the optimal control model of vehicle body stability at the matching points to obtain the processed variables, assign values ​​to the processed variables at the matching points, obtain the vehicle body posture, hydraulic cylinder angle and hydraulic cylinder speed time series trajectory, and calculate the objective function and constraint violation of the optimal control model of vehicle body stability under the current driving trajectory according to the vehicle body posture, hydraulic cylinder angle and hydraulic cylinder speed time series trajectory, including the hydraulic cylinder extension length, hydraulic cylinder extension speed, and the distance from the tire center to the ground, to determine whether the optimal control model of vehicle body stability meets the optimization convergence conditions. If so, the optimal control variables of the optimal control model of vehicle body stability are output. If not, the state variables and control variables of the optimal control model of vehicle body stability are optimized in a single-step iterative manner, and the processed variables at the matching points are reassigned until the optimal control variables of the optimal control model of vehicle body stability are output to achieve the stability control of the walking excavator body.

[0029] According to a third aspect of an embodiment of the present invention, there is provided an electronic device, comprising a processor and a memory storing a program, wherein the program comprises instructions, and when the instructions are executed by the processor, the processor executes the steps executed by the method of the first aspect.

[0030] According to a fourth aspect of an embodiment of the present invention, there is provided a computer storage medium on which a computer program is stored. When the program is executed by a processor, the method of the first aspect described above is implemented.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] The present invention constructs a mapping relationship between road excitation, the angle between the walking legs and the chassis, and the body posture, takes the walking leg angle as the optimization variable, and the pitch angle, roll angle and vertical displacement as the optimization targets. An intelligent optimization algorithm is used to achieve precise adjustment of the walking leg angle, thereby significantly improving the posture stability and passability of the crawler excavator in complex terrain, which is of great significance for improving work efficiency and quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the embodiments of the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0034] Figure 1 The present invention is a schematic diagram of a walking excavator applicable to the present invention.

[0035] Figure 2 It is a schematic diagram of terrain scanning in front of a vehicle by a laser radar of the present invention.

[0036] Figure 3 It is a schematic diagram of the excavator body dynamics parameters of the present invention.

[0037] Figure 4 The present invention is a flowchart of the steps of the method for stabilizing the posture of the vehicle body of an unmanned walking excavator based on a neural network.

[0038] Figure 5 For Figure 4 The corresponding overall flow chart of the unmanned walking excavator body posture stability control method based on neural network. DETAILED DESCRIPTION

[0039] In order to have a clearer understanding of the technical features, purposes and effects of the embodiments of the present invention, the specific implementation of the embodiments of the present invention is now described with reference to the accompanying drawings.

[0040] In this document, “exemplary” means “serving as an example, instance or illustration”, and any illustration or implementation described in this document as “exemplary” should not be construed as a more preferred or more advantageous technical solution.

[0041] In order to enable those skilled in the art to better understand the technical solutions in the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in the field based on the embodiments in the embodiments of the present invention should fall within the scope of protection of the embodiments of the present invention.

[0042] The specific implementation of the embodiment of the present invention is further described below in conjunction with the accompanying drawings of the embodiment of the present invention.

[0043] The method for controlling the posture stability of an unmanned walking excavator based on a neural network is applicable to Figure 1 The unmanned crawler excavator shown in the figure achieves stable control of the body posture of the excavator on complex road conditions by adjusting the walking leg hydraulic cylinders.

[0044] The present invention uses multiple sensors such as laser radar (LiDAR) and cameras to scan the terrain in front of the excavator, and collect multi-dimensional feature information such as height, slope and unevenness. After data registration and filtering processing, accurate three-dimensional point cloud data is obtained, and mathematical fitting methods such as least squares vector machine are used to resolve it into terrain excitation data containing slope, smoothness and stiffness; secondly, according to the neural network posture prediction modeling, the model takes terrain excitation and the angle between the walking leg and the chassis as input, and takes pitch angle, roll angle and vertical displacement as output. Through deep learning training of a large amount of operation data, the model can accurately predict the changes in body posture under different terrains and walking leg configurations, and construct a nonlinear mapping relationship between the walking leg angle and the body posture. Furthermore, the walking leg angle is used as the optimization variable, and the stability of the body posture (pitch angle, roll angle and vertical displacement) is used as the objective function. The model sets multiple constraints, including the range of the walking leg angle and its angular velocity limit to build the optimal control model for vehicle body stability. The optimal control model is transformed into a nonlinear programming problem through discretization method, and the optimal output force of the hydraulic cylinder and the optimal motion trajectory of the vehicle body are gradually iterated through the nonlinear optimization algorithm. Figure 2 As shown, the schematic diagram of the excavator body dynamics parameters of the present invention is as follows Figure 3 shown.

[0045] The present invention proposes a method for controlling the posture of an unmanned walking excavator based on a neural network. The control object is an ET120 walking excavator, and the terrain scanning device is a Hesai Pandar 64-line laser radar. The flowchart of the method proposed by the present invention is as follows: Figure 4 , Figure 5 As shown, the following steps are included:

[0046] S1, scanning the terrain in front of the walking excavator through a laser radar and a camera, generating terrain point cloud data, processing the terrain point cloud data, and obtaining terrain excitation data;

[0047] S2, taking the terrain excitation data and the walking leg angle of the walking excavator as input, constructing a neural network posture prediction model of the walking excavator, and outputting the pitch angle, roll angle and vertical displacement of the walking excavator body to predict the body posture change of the walking excavator;

[0048] S3. Based on the body posture change of the walking excavator, an optimal control model of the body stability of the walking excavator is constructed, including: optimization variables, objective functions, and model constraints;

[0049] S4, assigning values ​​to parameters in the vehicle body stability optimal control model;

[0050] S5, discretizing the variables of the vehicle body stability optimal control model at the distribution points to obtain processed variables;

[0051] S6, assigning values ​​to the processed variables at the matching points to obtain the vehicle body posture, hydraulic cylinder angle and hydraulic cylinder speed time series trajectory;

[0052] S7, calculating the objective function and constraint violation degree of the vehicle body stability optimal control model under the current driving trajectory according to the vehicle body posture, hydraulic cylinder angle and hydraulic cylinder speed time series trajectory, including the hydraulic cylinder extension length, hydraulic cylinder extension speed, and the distance from the tire center to the ground;

[0053] S8, judging whether the vehicle body stability optimal control model satisfies the optimization convergence condition, if so, outputting the optimal control variables of the vehicle body stability optimal control model, if not, turning to step S9;

[0054] S9, single-step iterative optimization of the state variables and control variables of the vehicle body stability optimal control model, and transfer to step S6, until the optimal control variables of the vehicle body stability optimal control model are output to achieve stability control of the walking excavator body.

[0055] Step S1 is divided into the following two steps:

[0056] (1) The height, slope, and bumps of the terrain in front of the excavator are collected through LiDAR and cameras. The LiDAR generates three-dimensional point cloud data, and the camera provides texture and color features. The data space is aligned using calibration and registration technology, and statistical filtering, radius filtering, denoising, Voxel Grid filtering, downsampling, and RANSAC algorithm are used to segment the plane to extract terrain features, namely terrain excitation data (such as slope and curvature).

[0057] The preprocessed terrain features are input into the least squares support vector machine (LS-SVR). Through kernel function mapping and optimization of the objective function, the relationship between terrain features and output targets (such as terrain type and slope) is learned to achieve terrain classification or characterization, providing support for excavator operations.

[0058] (2) Use the least squares support vector machine to establish a nonlinear regression relationship model between the x, y and elevation z of the terrain point:

[0059] z=f(x,y)=ω Τ φ(x,y)+b (1)

[0060] Among them, φ(x,y) is the feature mapping function, which maps low-dimensional features to high-dimensional space; ω is the regression coefficient, and b is the bias term. Modeling is performed by minimizing the following optimization objective function:

[0061]

[0062] The constraints are:

[0063] z i =ω Τ φ(x i ,y i )+b+ξ i ,i=1,2,...,N (3)

[0064] Among them, ξ i is the slack variable and γ is the regularization parameter.

[0065] The nonlinear regression model established based on LS-SVR can be substituted into formula (1) to calculate the road elevation information z at any horizontal coordinate (x, y) of the terrain in front of the vehicle.

[0066] Step S2, the modeling process of the neural network posture prediction model of the walking excavator is as follows:

[0067] (1) Based on the terrain excitation data obtained in step S1 (including information such as slope, smoothness, and convexity) and the operation data of the excavator (such as the walking leg angle and the chassis angle), a training data set for the neural network posture prediction model is constructed. The input data includes terrain excitation (w1, w2, w3, w4) and the angle between the walking leg and the chassis (θ1, θ2, θ3, θ4), and the output is the body posture parameter: pitch angle θ s , roll angle and vertical displacement z s .

[0068] (2) The posture prediction model adopts a multi-layer feedforward neural network (FNN) structure, including an input layer, several hidden layers, and an output layer. The number of nodes in the input layer is n+4, where n represents the number of terrain excitation features and the four nodes represent the angle between the walking leg and the chassis. The hidden layer is composed of multiple layers of neurons, and the activation function uses ReLU to enhance the fitting ability of complex nonlinear relationships. The output layer contains three nodes, corresponding to the pitch angle θ s , roll angle and vertical displacement z s .

[0069] (3) Define the loss function and optimization target, and train the network through supervised learning. Select the mean squared error (MSE) loss function to minimize the error between the network prediction output and the true posture data. The MSE loss function is defined as:

[0070]

[0071] in, Represents the output value predicted by the model, y i is the true posture value, and m is the number of samples.

[0072] The optimization method uses the Adam optimization algorithm to iteratively update the network weights according to the gradient information to accelerate convergence. The Adam algorithm dynamically adjusts the learning rate based on the average value of the gradient of the first-order and second-order moments, so that the network can reach the optimal solution faster.

[0073] (4) Through multiple rounds of iterative training of the neural network posture prediction model, the model continuously updates the weight parameters and gradually reduces the error between the predicted output and the true value. To prevent overfitting, regularization and early stopping strategies are adopted, and the optimal model structure and hyperparameters are determined through cross-validation. After training, the neural network can accurately predict the changes in vehicle posture based on the input terrain excitation and walking leg angle. By inputting specific terrain excitation and walking leg configuration, the model outputs the pitch angle, roll angle and vertical displacement of the vehicle body, and successfully establishes a nonlinear mapping relationship between walking leg configuration and vehicle posture.

[0074] Step S3, the construction of the optimal control model of the body stability of the walking excavator can be divided into the following three steps:

[0075] (1) Taking the walking leg angle as the optimization variable and the stability of the vehicle body posture as the objective function, the optimal control model of vehicle body stability is constructed. The objective function is expressed as:

[0076]

[0077] Among them, θ s represents the pitch angle, is the roll angle, z s represents the vertical displacement, t f Indicates the excavator operating time.

[0078] (2) In order to ensure that the excavation trajectory and input force of the walking excavator meet the actual feasibility during the movement, relevant geometric and performance constraints are made in the optimization process. The model constraints include the following aspects:

[0079] a During the posture adjustment process, the hydraulic cylinders of the four supporting legs of the walking excavator are extended and retracted at a speed v hi (i=1,2,3,4) to ensure that the maximum allowable lower limit v hmin and the maximum allowed upper limit v hmax between:

[0080]

[0081] b During the movement, the center of the four-leg tires of the walking excavator is displaced by z ui (i=1,2,3,4) and the roadside elevation w corresponding to each wheel i The difference is kept within the maximum permissible lower limit ε min and the maximum permissible upper limit ε max between:

[0082]

[0083] (3) Combining the above steps a and b, the optimal control model of the body stability of the walking excavator is expressed as:

[0084] findθ i (i=1,2,3,4)

[0085]

[0086] sC ij ≤0(i=1,2,3,4,j=1,2,3,4)

[0087] t min≤t f ≤t max

[0088] Among them, θ i (i=1,2,3,4) represents the walking leg angle, θ s represents the pitch angle, is the roll angle, z s represents the vertical displacement, t f represents the running time, t min represents the minimum running time, t max Indicates the maximum running time.

[0089] Step S5, the steps of discretizing the optimal control model of vehicle body stability are as follows:

[0090] In order to solve the optimization problem of the optimal control model of vehicle body stability, the integral term in the objective function is first discretized. f | is divided into N small intervals, and the integral is approximately in the form of summation:

[0091]

[0092] in, is the time length of each discrete step. Through discretization, the optimization problem is transformed into solving the optimal combination of a series of discrete variables.

[0093] Next, we use the Genetic Algorithm (GA) to solve the problem, including:

[0094] The walking leg angle θ i (i=1,2,3,4) is encoded as the gene of the individual, and the population is initialized so that each individual represents an angle combination;

[0095] By defining the objective function Calculate the fitness of each individual, and retain and generate excellent individuals in operations such as selection, crossover and mutation;

[0096] The penalty function method is used to reduce the fitness of individuals that do not meet the constraints, so as to meet the constraints of the angle range and angular velocity of the walking legs.

[0097] The optimal control model of vehicle body stability is iteratively evolved, and the optimal control variables of the optimal control model of vehicle body stability are output to obtain the optimal solution of the walking leg angle combination and realize the stability control of the walking excavator body.

[0098] The present invention introduces a neural network to establish a mapping relationship between road excitation, walking legs and vehicle body posture, constructs a prediction model for vehicle body posture under arbitrary terrain, and further establishes an optimization method for solving it. This is of great significance for ensuring the safety of unmanned crawler excavators during autonomous operation and improving the operation quality.

[0099] The embodiment of the present invention also provides a body posture stability control system for an unmanned walking excavator based on a neural network, comprising a road feature recognition module, i.e., a data acquisition module, a neural network posture prediction model module, a body stability optimal control model module, and a body stability optimal control model solution module, wherein:

[0100] A data acquisition module is used to scan the terrain in front of the walking excavator through a laser radar and a camera, generate terrain point cloud data, process the terrain point cloud data, and obtain terrain excitation data;

[0101] A neural network posture prediction model module is used to take terrain excitation data and the walking leg angle of the walking excavator as input, build a neural network posture prediction model of the walking excavator, and output the pitch angle, roll angle and vertical displacement of the walking excavator body to predict the body posture change of the walking excavator;

[0102] The vehicle body stability optimal control model module is used to construct the vehicle body stability optimal control model of the walking excavator based on the vehicle body posture change of the walking excavator, including: optimizing variables, objective functions, model constraints, and assigning parameters in the vehicle body stability optimal control model;

[0103] The vehicle body stability optimal control model solving module is used to discretize the variables of the vehicle body stability optimal control model at the matching points to obtain the processed variables, assign values ​​to the processed variables at the matching points, obtain the vehicle body posture, hydraulic cylinder angle and hydraulic cylinder speed time series trajectory, and calculate the objective function and constraint violation degree of the vehicle body stability optimal control model under the current driving trajectory according to the vehicle body posture, hydraulic cylinder angle and hydraulic cylinder speed time series trajectory, including the hydraulic cylinder extension length, hydraulic cylinder extension speed, and the distance from the tire center to the ground, to determine whether the vehicle body stability optimal control model meets the optimization convergence conditions. If so, the optimal control variables of the vehicle body stability optimal control model are output. If not, the state variables and control variables of the vehicle body stability optimal control model are optimized by single-step iterative optimization, and the processed variables at the matching points are reassigned until the optimal control variables of the vehicle body stability optimal control model are output to achieve the stability control of the walking excavator body.

[0104] The first part of the data acquisition module of the present invention uses multiple sensors such as laser radar and camera to scan the terrain in front of the excavator, and collects multi-dimensional feature information such as height, slope and convexity. After data registration and filtering, accurate three-dimensional point cloud data is obtained, and mathematical fitting methods such as least squares support vector machine are used to parse it into terrain excitation data including slope, smoothness and stiffness; the second part is a neural network posture prediction model module, which takes terrain excitation and the angle between the walking leg and the chassis as input, and takes pitch angle, roll angle and vertical displacement as output. Through deep learning training of a large amount of operation data, the model can accurately predict the changes in body posture under different terrains and walking leg configurations, and construct a nonlinear mapping relationship between the walking leg angle and the body posture; the third part is the body stability optimal control model module. The walking leg angle is used as the optimization variable, and the stability of the body posture (pitch angle, roll angle and vertical displacement) is used as the objective function. The model sets multiple constraints, including the range of the walking leg angle and its angular velocity limit; the fourth part is the body stability optimal control model solution module. The optimal control model is transformed into a nonlinear programming problem through a discretization method, and the optimal motion trajectory of the vehicle body is obtained step by step through the nonlinear optimization algorithm.

[0105] Compared with the prior art, the present invention has the following beneficial effects:

[0106] The present invention constructs a mapping relationship between road excitation, the angle between the walking legs and the chassis, and the body posture, takes the walking leg angle as the optimization variable, and the pitch angle, roll angle and vertical displacement as the optimization targets. An intelligent optimization algorithm is used to achieve precise adjustment of the walking leg angle, thereby significantly improving the posture stability and passability of the crawler excavator in complex terrain, which is of great significance for improving work efficiency and quality.

[0107] As another example, the present invention also provides an electronic device, and now will describe an electronic device that can be used as a server or client of the present invention, which is an example of a hardware device that can be applied to various aspects of the present invention. Electronic devices are intended to represent various forms of digital electronic computer devices, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic devices can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices and other similar computing devices. The components shown herein, their connections and relationships, and their functions are only examples, and are not intended to limit the implementation of the present invention described and / or required herein.

[0108] The electronic device may include: a processor (processor), a communication interface (CommunicationsInterface), a memory (memory) and a communication bus.

[0109] The processor, the communication interface and the memory communicate with each other through the communication bus. The communication interface is used to communicate with other electronic devices or servers.

[0110] The processor is used to execute the program, and specifically can execute the relevant steps in the above method embodiment.

[0111] Specifically, the program may include program codes including computer operation instructions.

[0112] The processor may be a CPU, or an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of the present invention. The one or more processors included in the smart device may be processors of the same type, such as one or more CPUs; or processors of different types, such as one or more CPUs and one or more ASICs.

[0113] The memory is used to store programs and may include a high-speed RAM memory and may also include a non-volatile memory, such as at least one disk memory.

[0114] When the program is executed by the processor, it is used to enable the electronic device to execute the neural network-based unmanned walking excavator body posture stabilization control method of the present invention.

[0115] In addition, the specific implementation of each step in the program can refer to the corresponding description of the corresponding steps and units in the above method embodiment, which will not be repeated here. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process of the above-described devices and modules can refer to the corresponding process description in the above method embodiment, which will not be repeated here.

[0116] The exemplary embodiments of the present invention further provide a computer storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the methods of the various embodiments of the present invention. The corresponding process descriptions in the aforementioned method embodiments may be referred to and will not be repeated here.

[0117] The above-described method according to an embodiment of the present invention may be implemented in hardware, firmware, or as software or computer code that may be stored in a recording medium (such as a CD ROM, RAM, floppy disk, hard disk, or magneto-optical disk), or as computer code that is originally stored in a remote recording medium or a non-temporary machine-readable medium downloaded over a network and will be stored in a local recording medium, so that the method described herein may be stored in such software processing on a recording medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware (such as an ASIC or FPGA). It is understood that a computer, processor, microprocessor controller, or programmable hardware includes a storage component (e.g., RAM, ROM, flash memory, etc.) that can store or receive software or computer code, and when the software or computer code is accessed and executed by a computer, processor, or hardware, the method described herein is implemented. In addition, when a general-purpose computer accesses the code for implementing the method shown herein, the execution of the code converts the general-purpose computer into a dedicated computer for executing the method shown herein.

[0118] Thus far, specific embodiments of the present invention have been described. Other embodiments are within the scope of the appended claims. In some cases, the actions recited in the claims may be performed in a different order and still achieve the desired results. Additionally, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In some embodiments, multitasking and parallel processing may be advantageous.

[0119] It should be understood that although this specification is described according to various embodiments, not every embodiment contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

[0120] Finally, it should be noted that the above implementation methods are only used to illustrate the embodiments of the present invention, and are not limitations of the embodiments of the present invention. Ordinary technicians in the relevant technical field can make various changes and modifications without departing from the spirit and scope of the embodiments of the present invention. Therefore, all equivalent technical solutions also belong to the scope of the embodiments of the present invention. The patent protection scope of the embodiments of the present invention should be defined by the claims.

Claims

1. A method for controlling the body posture of an unmanned walking excavator based on a neural network, characterized in that: The specific steps are as follows: S1, scanning the terrain in front of the walking excavator through a laser radar and a camera, generating terrain point cloud data, processing the terrain point cloud data, and obtaining terrain excitation data; S2, taking the terrain excitation data and the walking leg angle of the walking excavator as input, constructing a neural network posture prediction model of the walking excavator, and outputting the pitch angle, roll angle and vertical displacement of the walking excavator body to predict the body posture change of the walking excavator; S3. Based on the body posture change of the walking excavator, an optimal control model of the body stability of the walking excavator is constructed, including: optimization variables, objective functions, and model constraints; S4, assigning values ​​to parameters in the vehicle body stability optimal control model; S5, discretizing the variables of the vehicle body stability optimal control model at the distribution points to obtain processed variables; S6, assigning values ​​to the processed variables at the matching points to obtain the vehicle body posture, hydraulic cylinder angle and hydraulic cylinder speed time series trajectory; S7, calculating the objective function and constraint violation degree of the vehicle body stability optimal control model under the current driving trajectory according to the vehicle body posture, hydraulic cylinder angle and hydraulic cylinder speed time series trajectory, including the hydraulic cylinder extension length, hydraulic cylinder extension speed, and the distance from the tire center to the ground; S8, judging whether the vehicle body stability optimal control model satisfies the optimization convergence condition, if so, outputting the optimal control variables of the vehicle body stability optimal control model, if not, turning to step S9; S9, single-step iterative optimization of the state variables and control variables of the vehicle body stability optimal control model, and transfer to step S6, until the optimal control variables of the vehicle body stability optimal control model are output to achieve stability control of the walking excavator body.

2. The method according to claim 1, characterized in that The terrain excitation data in step S1 includes slope, smoothness and concavity.

3. The method according to claim 2, characterized in that In step S1, the terrain point cloud data is processed by at least one of statistical filtering, radius filtering denoising, VoxelGrid filtering downsampling, and RANSAC algorithm plane segmentation.

4. The method according to claim 3, characterized in that In step S2, the method for establishing the neural network posture prediction model of the walking excavator is as follows: Based on the terrain excitation data and the angle of the walking legs of the excavator, a training data set of the neural network posture prediction model is constructed. The input data of the neural network posture prediction model includes terrain excitation (w1, w2, w3, w4) and the angle between the walking legs and the chassis (θ1, θ2, θ3, θ4). The output data is the body posture parameters, including: pitch angle θ s , roll angle and vertical displacement z s ; The neural network posture prediction model adopts a multi-layer feedforward neural network structure, including an input layer, several hidden layers and an output layer. The number of nodes in the input layer is n+4, where n represents the number of terrain excitation features. The 4 nodes represent the angle between the walking leg and the chassis. The hidden layer is composed of multiple layers of neurons, and the activation function uses ReLU. The output layer contains 3 nodes, corresponding to the pitch angle θ s , roll angle and vertical displacement z s ; Define the loss function and optimization target, train the neural network posture prediction model through supervised learning, select the mean square error loss function to minimize the error between the predicted output of the neural network posture prediction model and the actual posture data. The mean square error loss function is defined as: in, Represents the output value predicted by the neural network posture prediction model, y i is the true posture value, m is the number of samples; The optimization method uses the Adam optimization algorithm to iteratively update the weights of the neural network posture prediction model according to the gradient information to accelerate convergence; Through multiple rounds of iterative training of the neural network posture prediction model, the weight parameters are updated to reduce the error between the predicted output and the true value; Regularization and early stopping strategies are adopted, and the optimal model structure and hyperparameters of the neural network posture prediction model are determined through cross-validation to prevent the neural network posture prediction model from overfitting; After training, the neural network posture prediction model can accurately predict the changes in the body posture of the walking excavator based on the input terrain excitation data and the walking leg angle.

5. The method according to claim 4, characterized in that The Adam optimization algorithm dynamically adjusts the learning rate based on the gradient average of the first-order and second-order moments to speed up the neural network posture prediction model to reach the optimal solution.

6. The method according to claim 1, characterized in that In step S3, the steps of constructing the optimal control model of the vehicle body stability of the walking excavator are as follows: Taking the walking leg angle as the optimization variable and the stability of the vehicle body posture as the objective function, the optimal control model of vehicle body stability is constructed. The objective function is expressed as: Among them, θ s represents the pitch angle, is the roll angle, z s represents the vertical displacement, t f Indicates the excavator operation time; Model constraints include: During the posture adjustment process, the four support leg hydraulic cylinders of the walking excavator are extended and retracted at a speed v hi (i=1,2,3,4) to ensure that the maximum allowable lower limit v hmin and the maximum allowed upper limit v hmax between: During the movement, the center of the four-leg tires of the walking excavator is displaced by z ui (i=1,2,3,4) and the roadside elevation w corresponding to each wheel i The difference is kept within the maximum permissible lower limit ε min and the maximum permissible upper limit ε max between: Therefore, the optimal control model of the body stability of the walking excavator is expressed as: findθ i (i=1,2,3,4) s.t.C ij ≤0(i=1,2,3,4,j=1,2,3,4) t min ≤t f ≤t max Among them, θ i (i=1,2,3,4) represents the walking leg angle, θ s represents the pitch angle, is the roll angle, z s represents the vertical displacement, t f represents the running time, t min represents the minimum running time, t max Indicates the maximum running time.

7. The method according to claim 4, characterized in that In step S5, the steps of discretizing the optimal control model of vehicle body stability are as follows: First, the integral term in the objective function is discretized and the time interval |0,t f | is divided into N small intervals, and the integral is approximately in the form of summation: in, For the time length of each discrete step, the optimization problem of the vehicle body stability optimal control model is transformed into solving the optimal combination of discrete variables through discretization; Then the genetic algorithm is used to solve the problem, including: The walking leg angle θ i (i=1,2,3,4) is encoded as the gene of the individual, and the population is initialized so that each individual represents an angle combination; By defining the objective function Calculate the fitness of each individual, and retain and generate excellent individuals in selection, crossover and mutation operations; The penalty function method is used to reduce the fitness of individuals that do not meet the constraints, so as to meet the constraints of the angle range and angular velocity of the walking legs. The optimal control model of vehicle body stability is iteratively evolved to obtain the optimal solution of walking leg angle combination and realize the stability control of the walking excavator body.

8. A body posture stabilization control system for an unmanned walking excavator based on a neural network, characterized in that: include: A data acquisition module is used to scan the terrain in front of the walking excavator through a laser radar and a camera, generate terrain point cloud data, process the terrain point cloud data, and obtain terrain excitation data; A neural network posture prediction model module is used to take terrain excitation data and the walking leg angle of the walking excavator as input, build a neural network posture prediction model of the walking excavator, and output the pitch angle, roll angle and vertical displacement of the walking excavator body to predict the body posture change of the walking excavator; The vehicle body stability optimal control model module is used to construct the vehicle body stability optimal control model of the walking excavator based on the vehicle body posture change of the walking excavator, including: optimizing variables, objective functions, model constraints, and assigning parameters in the vehicle body stability optimal control model; The vehicle body stability optimal control model solving module is used to discretize the variables of the vehicle body stability optimal control model at the matching points to obtain the processed variables, assign values ​​to the processed variables at the matching points, obtain the vehicle body posture, hydraulic cylinder angle and hydraulic cylinder speed time series trajectory, and calculate the objective function and constraint violation degree of the vehicle body stability optimal control model under the current driving trajectory according to the vehicle body posture, hydraulic cylinder angle and hydraulic cylinder speed time series trajectory, including the hydraulic cylinder extension length, hydraulic cylinder extension speed, and the distance from the tire center to the ground, to determine whether the vehicle body stability optimal control model meets the optimization convergence conditions. If so, the optimal control variables of the vehicle body stability optimal control model are output. If not, the state variables and control variables of the vehicle body stability optimal control model are optimized by single-step iterative optimization, and the processed variables at the matching points are reassigned until the optimal control variables of the vehicle body stability optimal control model are output to achieve the stability control of the walking excavator body.

9. An electronic device, characterized in that: include: processor; A memory for storing programs; The program includes instructions, which, when executed by the processor, cause the processor to perform the steps of the method as claimed in any one of claims 1 to 7.

10. A computer storage medium, characterized in that: A computer program is stored thereon, and when the program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images