A method and system for predicting the working posture of intelligent engineering machinery

Abstract

The present invention discloses a method and system for predicting the working posture of intelligent engineering machinery, the method comprising the following steps: obtaining the current working posture information of the engineering machinery; based on the current working posture information of the engineering machinery, using an intelligent extended Kalman filter algorithm to predict the next working posture of the engineering machinery; the present invention first uses the salp algorithm to search for the best, when the salp algorithm falls into a local optimal state, the particle swarm algorithm is integrated into the salp algorithm and the search space is increased, and the improved salp algorithm is obtained by multiple cycles. The improved salp algorithm is used to optimize the system noise covariance matrix in the extended Kalman filter algorithm to obtain the intelligent extended Kalman filter algorithm. Based on the posture information obtained by the inertial measurement unit measurement analysis, the intelligent extended Kalman filter algorithm is used to predict the next action of the engineering machinery, thereby improving the accuracy of the prediction results of the working posture of the engineering machinery.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent operation of engineering machinery, and in particular to a method and system for predicting the operation posture of intelligent engineering machinery. Background Art

[0002] Excavators or backhoe loaders are common, powerful construction machines widely used in a variety of fields, including disaster relief and infrastructure construction. However, operating in complex conditions requires highly skilled operators, which can be lengthy, costly, and ineffective. Furthermore, excavators and backhoe loaders often operate in environments subject to noise, dust, and even harsh environments that can endanger the operators' lives.

[0003] In recent years, unmanned intelligent operation of construction machinery has become a hot topic among researchers, driven by the need to address these challenges. Accurately acquiring and predicting the machine's operating posture is paramount to achieving this goal. The extended Kalman filter is a commonly used posture prediction method, but the selection of its process and measurement noise matrices can introduce errors in posture prediction, impacting the accuracy of intelligent operation. Therefore, current research focuses on using intelligent algorithms to improve these two matrices. Summary of the Invention

[0004] In response to the problems existing in the prior art, the present invention provides a method and system for predicting the operating posture of intelligent engineering machinery. The method adopts the Salp Albumin algorithm for optimization. When the Salp Albumin algorithm falls into a local optimal state, the particle swarm algorithm is integrated into the Salp Albumin algorithm and the search space is increased. An improved Salp Albumin algorithm is obtained through multiple cycles. The improved Salp Albumin algorithm is used to optimize the system noise covariance matrix in the extended Kalman filter algorithm to obtain an intelligent extended Kalman filter algorithm. Based on the posture information obtained by measurement and analysis of the inertial measurement unit, the intelligent extended Kalman filter algorithm is used to predict the next action of the engineering machinery, thereby improving the accuracy of the prediction results of the operating posture of the engineering machinery.

[0005] The technical solutions of the present invention are as follows:

[0006] In a first aspect of the present invention, a method for predicting the working posture of an intelligent engineering machine is provided, comprising the following steps:

[0007] Obtain the current operating posture information of the construction machinery;

[0008] Based on the current operating posture information of the construction machinery, an intelligent extended Kalman filter algorithm is used to predict the next operating posture of the construction machinery. Specifically, the salp algorithm is used for optimization. When the salp algorithm falls into a local optimal state, the particle swarm optimization algorithm is integrated into the salp algorithm and the search space is increased. After multiple cycles, an improved salp algorithm is obtained. The improved salp algorithm is used to optimize the system noise covariance matrix in the extended Kalman filter algorithm to obtain the intelligent extended Kalman filter algorithm. The intelligent extended Kalman filter algorithm is used to predict the engineering operation posture.

[0009] In some embodiments of the present invention, the engineering machinery is an excavator or an excavator loader. During the specific operation process, the coordinated movement of its boom, dipper arm, and bucket completes the excavation operation; during the excavation operation, the operation process is: the engineering machinery moves to a suitable position, the bucket extends forward, the boom is lowered until the bucket touches the ground, and then the dipper arm is manipulated to make the bucket complete the excavation and loading work. After the bucket is full, it is raised together with the boom, and then the bucket is rotated to a suitable unloading position to unload the soil. After unloading is completed, it is turned to a suitable excavation position again to perform a second excavation cycle.

[0010] In some embodiments of the present invention, the inertial measurement includes a gyroscope for measuring three-axis angular velocity, an accelerometer for measuring three-axis acceleration, and a magnetometer for providing three-axis orientation information; the inertial measurement device is fixedly connected to the excavator or backhoe loader motion carrier, and after coordinate transformation of each coordinate system, the angular velocity information measured by the gyroscope and the acceleration information measured by the accelerometer are converted to the global coordinate system, and the influence of gravity acceleration is removed; a quaternion attitude update method is adopted in the inertial measurement unit, which uses a high-order complex form for coordinate transformation, and obtains the speed information and position information of the excavator or backhoe loader motion carrier in the global coordinate system through attitude analysis.

[0011] In some embodiments of the present invention, the salp algorithm assumes that the search space is an N×D Euclidean space, where N is the number of populations, D is the dimension of the space, and the position of the i-th individual in the space is: X i (0) = lb + (ub - lb) rand, where i = 1, 2, ..., N, indicating that the current individual is the i-th, lb is the lower limit vector of the search space, ub is the upper limit vector of the search space, and rand is a random number between 0 and 1. In the algorithm, N individuals are divided into two groups, namely leaders and followers. The leader is the first individual in the population, and its update formula is:

[0012]

[0013] Where, represents the position of the first salp in the d dimension, the leader, after t and t+1 iterations, F td is the location of the food in the d dimension, which is the location of the individual with the best fitness value, ub d lb d are the upper and lower boundaries of the d-dimensional space, is a random number between 0 and 1, is the convergence factor, and its expression is: Where t is the number of iterations and T is the maximum number of iterations;

[0014] Except for the leader, the rest of the salps in the population are followers. They move forward in a chain, and their displacement conforms to Newton's laws of motion. The motion displacement formula is: Where a is acceleration, and the calculation formula is: a=(v fianl -v0) / T t , v0 is the initial velocity, T t is the time; in the iterative process T t =1, v0=0, and the final formula for the follower's position is:

[0015]

[0016] In some embodiments of the present invention, the particle swarm algorithm assumes that the search space is D-dimensional and has N particles, then the particle swarm is represented as: X = {x1, x2, ... x N}, where X represents the particle swarm, x i represents the i-th particle in the population, then the position of the i-th particle can be expressed as: a i =(a1,a2,…,a D ) T , the corresponding velocity of the i-th particle is: v i =(v1,v2,…,v D ) T , the local optimal solution P of the i-th particle best For: P i =(P1,P2,…,P D ) T , the global optimal solution G of the entire particle swarm best For: G i =(G1,G2,…,G D ) T During the iteration process, the particle updates its position and velocity according to the current local optimal solution and the global optimal solution. The velocity update formula is:

[0017] v i (t+1)=ω·v i (t)+c1r1(P i (t)-a i(t))+c2r2(G i (t)-a i (t)),

[0018] The position update formula is:

[0019] a i (t+1)=a i (t)+v i (t+1),

[0020] Where t represents the tth iteration of the particle, P i (t) represents the local optimum of the particle after the tth update, G i (t) represents the global optimum of the particle after the tth update, v i (t+1) represents the velocity of particle i after t+1 iterations, a i (t+1) represents the position of particle i after t+1 iterations, c1 and c2 are learning factors, namely individual learning factor and social learning factor, r1 and r2 are random numbers between 0 and 1, which are used to represent the randomness of each particle in the search process of the particle swarm algorithm to increase the possibility of reaching the optimal solution. The inertia coefficient ω is used to balance the local search and global search.

[0021] In some embodiments of the present invention, the particle swarm algorithm is integrated into the salp algorithm, including: introducing a velocity update formula and a position update formula of the particle swarm algorithm to improve the leader and follower update formulas in the salp algorithm;

[0022] The leader update formula after the improved Salp Algorithm is:

[0023]

[0024] Where, They represent the position of the first salp in the d dimension, i.e., the leader, after t and t+1 iterations, respectively. t d is the location of the food in the d dimension, which is the location of the individual with the best fitness value, ub d lb d are the upper and lower boundaries of the d-dimensional space, is a random number between 0 and 1, is the convergence factor, and its expression is: Where t is the number of iterations, T is the maximum number of iterations, a and b are the weight coefficients of the salp algorithm and the particle swarm algorithm respectively. And a+b=1; V1 d (t+1) represents the speed of the leader in the d dimension at the t+1th iteration, and its expression is: Where ω is the inertia weight, r1 and r2 are two random numbers, ε1 and ε2 are learning factors, is the individual optimal solution, is the optimal solution for the population;

[0025] After the improvement of the Salp Algorithm, the update formula of the follower becomes:

[0026]

[0027] Where, Respectively represent the position of the i-th salp in the d dimension after t and t+1 iterations, V i d (t+1) represents the velocity of the i-th follower in the d-dimensional space at the t+1th iteration. ω is the inertia weight, r1 and r2 are two random numbers, ε1 and ε2 are learning factors, is the individual optimal solution, is the optimal solution for the population, T t is the time, which is 1 during the iteration process.

[0028] In some embodiments of the present invention, the method of increasing the search space and obtaining the improved salp algorithm through multiple cycles includes: firstly using the salp algorithm to find the optimal solution, and then It represents the difference between the food position after t+1 iterations and the food position after t iterations in the d dimension divided by the food position after t iterations; when the salp algorithm has λ(t+1)<1% for 20 consecutive times, the iterative change is very small, and the algorithm is randomly perturbed, that is, the salp algorithm fused with the particle swarm algorithm is used for optimization; then the search space of the salp algorithm is increased, and the upper limit of the d-dimensional space is set to Where, is the upper limit of the expanded d-dimensional space, ub d is the upper limit before the d-dimensional space is expanded, t is the number of iterations, and T is the maximum number of iterations; continue to search for the optimal solution, compare the optimal solution obtained by the search with the optimal solution obtained by the salp algorithm to obtain a new optimal solution, and use N zhq As the cumulative number, when N zhq When ≥3 or t=T reaches the maximum number of iterations, the iteration stops and the final optimal solution is obtained, and the improved salp algorithm is obtained.

[0029] In some embodiments of the present invention, the extended Kalman filter algorithm optimizes the nonlinear system using a method of Taylor expansion and retaining only the first-order Taylor series expansion term, and then estimates the state information of the system through the system's state equation and observation equation, uses the previous optimal result to predict the current value, and uses the observed value to correct the current value to obtain the optimal result; the system's state equation and measurement equation are:

[0030]

[0031] Where, the nonlinear functions f(·) and h(·) reflect the state vector and measurement vector The mapping relationship between k-1 is the process noise, v k is the measurement noise, which are independent of each other and obey N(0,Q k ) and N(0,R k )’s Gaussian white noise;

[0032] The basic formula of the extended Kalman filter algorithm is:

[0033] Time prediction equation

[0034] Where, represents the prior estimate of time k, and its corresponding prior error covariance matrix is represents the optimal estimate at time k-1, and its corresponding covariance matrix is μ k-1 is the system input at time k-1, Q k-1 is the covariance matrix of the system process noise at time k-1, A k For the equation The Jacobian matrix of ;

[0035] State update equation

[0036] Where K k is the Kalman gain matrix, H k For the equation The Jacobian matrix, R k represents the covariance matrix of the measurement noise at time k, Represents the optimal estimate at time k, and its corresponding covariance matrix is ​​P Pk .

[0037] In some embodiments of the present invention, the method of optimizing the system noise covariance matrix in the extended Kalman filter algorithm using the improved salp algorithm to obtain an intelligent extended Kalman filter algorithm includes: optimizing the system noise covariance matrix using the improved salp algorithm to obtain a modified covariance matrix as a new system noise covariance matrix in the extended Kalman filter, and using the extended Kalman filter to perform parameter estimation, and finally obtaining an optimized state estimate;

[0038] Furthermore, the use of the intelligent extended Kalman filter algorithm to predict the engineering operation posture includes: solving the angular velocity information and acceleration information measured by the inertial measurement unit to obtain the working posture of the engineering machinery, inputting the working posture of the engineering machinery into the intelligent extended Kalman filter algorithm to predict the working posture of the engineering machinery.

[0039] In a second aspect of the present invention, a system for predicting the working posture of an intelligent engineering machine is provided, which is characterized by comprising:

[0040] The operation posture information acquisition module is configured to: acquire the current operation posture information of the construction machinery;

[0041] The operation posture information prediction module is configured as follows: based on the current operation posture information of the engineering machinery, an intelligent extended Kalman filter algorithm is used to predict the next operation posture of the engineering machinery. Specifically, the salp algorithm is used for optimization. When the salp algorithm falls into a local optimal state, the particle swarm algorithm is integrated into the salp algorithm and the search space is increased. The improved salp algorithm is obtained through multiple cycles. The improved salp algorithm is used to optimize the system noise covariance matrix in the extended Kalman filter algorithm to obtain the intelligent extended Kalman filter algorithm. The intelligent extended Kalman filter algorithm is used to predict the engineering operation posture.

[0042] One or more technical solutions of the present invention have the following beneficial effects:

[0043] (1) The present invention improves the Salp Algorithm through the particle swarm algorithm, and adopts the Salp Algorithm for optimization. When the Salp Algorithm falls into a local optimal state, the particle swarm algorithm is integrated into the Salp Algorithm and the search space is increased. The improved Salp Algorithm is obtained through multiple cycles, thereby improving the global optimization ability of the Salp Algorithm.

[0044] (2) The present invention utilizes the improved Salp Intestine algorithm to optimize the system noise covariance matrix in the extended Kalman filter algorithm to obtain an intelligent extended Kalman filter algorithm. The improved Salp Intestine algorithm is used to optimize the system noise covariance matrix to obtain a modified covariance matrix, and the extended Kalman filter is used for parameter estimation to finally obtain an optimized state estimation value, thereby improving the accuracy of the engineering machinery operation posture prediction results.

[0045] (3) The method for predicting the operating posture of intelligent engineering machinery provided by the present invention is applicable to intelligent engineering machinery such as excavators or backhoe loaders, and can accurately predict the operating posture of engineering machinery, thereby improving the accuracy of unmanned intelligent operation of engineering machinery.

[0046] Advantages of additional aspects of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0048] Figure 1 This is a flow chart of the method for predicting the working posture of intelligent engineering machinery according to the present invention. DETAILED DESCRIPTION

[0049] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0050] Example 1

[0051] In a typical embodiment of the present invention, a method for predicting the working posture of intelligent engineering machinery is proposed. Figure 1 As shown, the following steps are included:

[0052] Step 1: Obtain the current working posture information of the construction machinery;

[0053] Step 2: Based on the current operating posture information of the construction machinery, the intelligent extended Kalman filter algorithm is used to predict the next operating posture of the construction machinery. Specifically, the salp algorithm is used for optimization. When the salp algorithm falls into a local optimal state, the particle swarm algorithm is integrated into the salp algorithm and the search space is increased. After multiple cycles, an improved salp algorithm is obtained. The improved salp algorithm is used to optimize the system noise covariance matrix in the extended Kalman filter algorithm to obtain the intelligent extended Kalman filter algorithm. The intelligent extended Kalman filter algorithm is used to predict the engineering operation posture.

[0054] In this embodiment, the construction machinery in step 1 can be an intelligent construction machine such as an excavator or backhoe loader. When performing excavation operations, an excavator or backhoe loader requires the coordinated movement of its boom, dipper arm, and bucket. The operation process is as follows: the excavator or backhoe loader moves to the appropriate position, the bucket extends forward, the boom lowers until the bucket touches the ground, and then the dipper arm is manipulated to complete the excavation and loading process. Once the bucket is full, it is raised along with the boom. The bucket then rotates to the appropriate unloading position to unload the soil. After unloading is complete, the bucket is returned to the appropriate digging position for a second digging cycle.

[0055] In this embodiment, the inertial measurement unit is a device capable of measuring the attitude angle and acceleration of an object, including a gyroscope for measuring three-axis angular velocity, an accelerometer for measuring three-axis acceleration, and a magnetometer that provides three-axis orientation information. The inertial measurement device is fixedly connected to the excavator or backhoe loader motion carrier. After the coordinate transformation of each coordinate system, the angular velocity information measured by the gyroscope and the acceleration information measured by the accelerometer can be converted to the global coordinate system, and the influence of gravity acceleration is removed. The quaternion attitude update method is adopted in the inertial measurement unit, which uses a high-order complex form for coordinate transformation. After attitude analysis, the speed information and position information of the excavator or backhoe loader motion carrier in the global coordinate system can be obtained.

[0056] In this embodiment, it is assumed that the search space of the Salp algorithm is an N×D Euclidean space, where N is the number of populations, D is the dimension of the space, and the position of the i-th individual in the space is: X i (0) = lb + (ub - lb) rand, where i = 1, 2, ..., N, indicating the current individual is the i-th, lb is the lower bound vector of the search space, ub is the upper bound vector of the search space, and rand is a random number between 0 and 1. The algorithm divides N individuals into two groups, leaders and followers. The leader is the first individual in the population, and its update formula is:

[0057]

[0058] Where, represents the position of the first salp (leader) in dimension d after t and t+1 iterations, F t d is the location of the food in the d dimension, which is the location of the individual with the best fitness value, ub d lb d are the upper and lower boundaries of the d-dimensional space, is a random number between 0 and 1, is the convergence factor, and its expression is: Where t is the number of iterations and T is the maximum number of iterations.

[0059] Except for the leader, the rest of the salps in the population are followers. They move forward in a chain, and their displacement conforms to Newton's laws of motion. The motion displacement formula is: Where a is acceleration, and the calculation formula is: a=(v fianl -v0) / T t , v0 is the initial velocity, T t is the time. In the iterative process T t =1, v0=0. The final formula for the follower's position is:

[0060]

[0061] In this embodiment, in the particle swarm algorithm, it is assumed that the search space is D-dimensional and there are N particles, then the particle swarm can be expressed as: X = {x1, x2, ... x N}, where X represents the particle swarm, x i represents the i-th particle in the population, then the position of the i-th particle can be expressed as: a i =(a1,a2,…,a D ) T , the corresponding velocity of the i-th particle is: v i =(v1,v2,…,v D ) T , the local optimal solution P of the i-th particle best For: P i =(P1,P2,…,P D ) T , the global optimal solution G of the entire particle swarm best For: G i =(G1,G2,…,G D ) T During the iteration process, the particle updates its position and velocity according to the current local optimal solution and the global optimal solution. The velocity update formula is:

[0062] v i (t+1)=ω·v i (t)+c1r1(P i (t)-a i (t))+c2r2(G i (t)-a i (t))

[0063] The position update formula is:

[0064] a i (t+1)=a i (t)+v i (t+1)

[0065] Where t represents the tth iteration of the particle, P i (t) represents the local optimum of the particle after the tth update, G i (t) represents the global optimum of the particle after the tth update, v i (t+1) represents the velocity of particle i after t+1 iterations, a i(t+1) represents the position of particle i after t+1 iterations, c1 and c2 are learning factors, namely individual learning factor and social learning factor, r1 and r2 are random numbers between 0 and 1, which are used to represent the randomness of each particle in the search process of the particle swarm algorithm to increase the possibility of reaching the optimal solution. The inertia coefficient ω is used to balance the local search and global search.

[0066] The particle swarm algorithm is integrated into the salp algorithm, and the speed update formula and position update formula of the particle swarm algorithm are used to improve the update formula of the leader and follower in the salp algorithm.

[0067] The leader update formula after the improved Salp Algorithm is:

[0068]

[0069] Where, represent the position of the first salp (leader) in the d dimension after t and t+1 iterations, respectively, and F t d is the location of the food in the d dimension, which is the location of the individual with the best fitness value, ub d lb d are the upper and lower boundaries of the d-dimensional space, is a random number between 0 and 1, is the convergence factor, and its expression is: Where t is the number of iterations, T is the maximum number of iterations, a and b are the weight coefficients of the salp algorithm and the particle swarm algorithm respectively. And a+b=1. V1 d (t+1) represents the speed of the leader in the d dimension at the t+1th iteration, and its expression is: Where ω is the inertia weight, r1 and r2 are two random numbers, ε1 and ε2 are learning factors, is the individual optimal solution, is the optimal solution for the population.

[0070] The follower update formula of the improved Salp Algorithm becomes:

[0071]

[0072] Where, Respectively represent the position of the i-th salp in the d dimension after t and t+1 iterations, V i d (t+1) represents the velocity of the i-th follower in the d-dimensional space at the t+1th iteration. ω is the inertia weight, r1 and r2 are two random numbers, ε1 and ε2 are learning factors, is the individual optimal solution, is the optimal solution for the population, Tt is the time, which is 1 during the iteration process.

[0073] The improved salp algorithm first uses the salp algorithm to find the optimal solution. It represents the difference between the food position after t+1 iterations and the food position after t iterations in the d dimension divided by the food position after t iterations. When the salp algorithm has λ(t+1) < 1% for 20 consecutive times, the iterative change is very small, and the algorithm is randomly perturbed, that is, the salp algorithm fused with the particle swarm algorithm is used for optimization. Then the search space of the salp algorithm is increased, and the upper limit of the d-dimensional space is set to Where, is the upper limit of the expanded d-dimensional space, ub d is the upper limit before the d-dimensional space is expanded, t is the number of iterations, and T is the maximum number of iterations. Continue to search for the optimal solution, compare the optimal solution obtained by the search with the optimal solution obtained by the Salp Sea Squid algorithm to obtain a new optimal solution, and use N zhq As the cumulative number, when N zhq When ≥3 or t=T reaches the maximum number of iterations, the iteration stops and the final optimal solution is obtained, thereby obtaining the improved salp algorithm.

[0074] The extended Kalman filter is a commonly used method for attitude prediction. It optimizes nonlinear systems by using Taylor expansion and retaining only the first-order Taylor series expansion term. It then estimates the state information of the system through the system's state equation and observation equation, uses the previous optimal result to predict the current value, and uses the observed value to correct the current value to obtain the optimal result. The state equation and measurement equation of the system are:

[0075]

[0076] Where, the nonlinear functions f(·) and h(·) reflect the state vector and measurement vector The mapping relationship between them. k-1 is the process noise, v k is the measurement noise, which are independent of each other and obey N(0,Q k ) and N(0,R k ) is Gaussian white noise.

[0077] The basic formula of the extended Kalman filter algorithm is:

[0078] Time prediction equation

[0079] Where, represents the prior estimate of time k, and its corresponding prior error covariance matrix is represents the optimal estimate at time k-1, and its corresponding covariance matrix is μ k-1 is the system input at time k-1, Q k-1 is the covariance matrix of the system process noise at time k-1, A k For the equation The Jacobian matrix of .

[0080] State update equation

[0081] Where K k is the Kalman gain matrix, H k For the equation The Jacobian matrix, R k represents the covariance matrix of the measurement noise at time k, Represents the optimal estimate at time k, and its corresponding covariance matrix is

[0082] The choice of noise in the extended Kalman filter algorithm will have a certain impact on the results of attitude prediction. Therefore, an intelligent algorithm is needed to improve it. The improved Salp Insipid algorithm is used to optimize the system noise covariance matrix Q to obtain a corrected covariance matrix, which is used as the new system noise covariance matrix in the extended Kalman filter. The extended Kalman filter is then used for parameter estimation to finally obtain the optimized state estimate.

[0083] Based on the posture information obtained by inertial measurement unit measurement and analysis, the intelligent extended Kalman filter algorithm obtained by optimizing the improved Salp Intestine algorithm is used to predict the next action of the construction machinery, thereby realizing the prediction of the operating posture of the intelligent construction machinery.

[0084] Example 2

[0085] In a second aspect of the present invention, a system for predicting the working posture of an intelligent engineering machine is provided, comprising:

[0086] The operation posture information acquisition module is configured to: acquire the current operation posture information of the construction machinery;

[0087] The operation posture information prediction module is configured as follows: based on the current operation posture information of the engineering machinery, an intelligent extended Kalman filter algorithm is used to predict the next operation posture of the engineering machinery. Specifically, the salp algorithm is used for optimization. When the salp algorithm falls into a local optimal state, the particle swarm algorithm is integrated into the salp algorithm and the search space is increased. The improved salp algorithm is obtained through multiple cycles. The improved salp algorithm is used to optimize the system noise covariance matrix in the extended Kalman filter algorithm to obtain the intelligent extended Kalman filter algorithm. The intelligent extended Kalman filter algorithm is used to predict the engineering operation posture.

[0088] Those skilled in the art will appreciate that the modules or steps of the present invention described above can be implemented using a general-purpose computer device. Alternatively, they can be implemented using program code executable by a computing device, which can then be stored in a storage device and executed by the computing device. Alternatively, they can be fabricated into separate integrated circuit modules, or multiple modules or steps can be fabricated into a single integrated circuit module for implementation. The present invention is not limited to any specific combination of hardware and software.

[0089] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A method for predicting the working posture of intelligent engineering machinery, characterized in that: The following steps are involved: Obtain the current working posture information of the construction machinery; Based on the current working posture information of the construction machinery, the intelligent extended Kalman filter algorithm is used to predict the next working posture of the construction machinery. Specifically, the salp algorithm is used for optimization. When the salp algorithm falls into a local optimal state, the particle swarm algorithm is integrated into the salp algorithm and the search space is increased. The improved salp algorithm is obtained by multiple cycles. The improved salp algorithm is used to optimize the system noise covariance matrix in the extended Kalman filter algorithm to obtain the intelligent extended Kalman filter algorithm. The intelligent extended Kalman filter algorithm is used to predict the working posture of the construction machinery. The particle swarm algorithm is integrated into the salp algorithm, including: introducing the speed update formula and position update formula of the particle swarm algorithm to improve the update formula of the leader and the follower in the salp algorithm; The update formula of the leader after the improvement of the Salp Algorithm is: In the formula, They represent the position of the first salp in the d dimension, i.e. the leader, after t and t+1 iterations, respectively. t d is the location of the food in the d dimension, which is the location of the individual with the best fitness value, ub d lb d are the upper and lower boundaries of the d-dimensional space, is a random number between 0 and 1, c c1 is the convergence factor, and its expression is: Where t is the number of iterations, T is the maximum number of iterations, a and b are the weight coefficients of the Salp Algorithm and the Particle Swarm Optimization Algorithm, respectively. And a+b=1; V1 d (t+1) represents the speed of the leader in the d dimension at the t+1th iteration, and its expression is: Where ω is the inertia weight, r1 and r2 are two random numbers, ε1 and ε2 are learning factors, is the individual optimal solution, is the optimal solution for the population; After the improvement of the Salp Algorithm, the update formula of the follower becomes: In the formula, denote the position of the i-th salp in d dimension after t and t+1 iterations, respectively. represents the speed of the ith follower in the d-dimensional space at the t+1th iteration, ω is the inertia weight, r1 and r2 are two random numbers, ε1 and ε2 are learning factors, is the individual optimal solution, is the optimal solution for the population, T t is the time, which is 1 during the iteration.

2. The method for predicting the working posture of intelligent engineering machinery according to claim 1, characterized in that: The engineering machinery is an excavator or an excavator loader. During the specific operation process, the coordinated movement of its boom, dipper arm and bucket completes the excavation operation. The operation process during the excavation operation is: the engineering machinery moves to a suitable position, the bucket extends forward, the boom is lowered until the bucket touches the ground, and then the dipper arm is manipulated to make the bucket complete the excavation and loading work. After the bucket is full, it is raised together with the boom, and then the bucket is rotated to a suitable unloading position to unload the soil. After the unloading of the soil is completed, it is turned to a suitable excavation position again to perform a second excavation cycle.

3. The method for predicting the working posture of intelligent engineering machinery according to claim 1, characterized in that: The inertial measurement unit includes a gyroscope for measuring three-axis angular velocity, an accelerometer for measuring three-axis acceleration, and a magnetometer for providing three-axis orientation information; the inertial measurement device is fixedly connected to the excavator or backhoe loader motion carrier, and after the coordinate transformation of each coordinate system, the angular velocity information measured by the gyroscope and the acceleration information measured by the accelerometer are converted to the global coordinate system, and the influence of gravity acceleration is removed; the quaternion attitude update method is adopted in the inertial measurement unit, which uses a high-order complex form to perform coordinate transformation, and obtains the speed information and position information of the excavator or backhoe loader motion carrier in the global coordinate system through attitude analysis.

4. The method for predicting the working posture of intelligent engineering machinery according to claim 1, characterized in that: The salp algorithm assumes that the search space is an N×D Euclidean space, where N is the population size, D is the space dimension, and the position of the i-th individual in the space is: i (0) = lb + (ub - lb) rand, where i = 1, 2, ..., N, indicating that the current individual is the i-th individual, lb is the lower limit vector of the search space, ub is the upper limit vector of the search space, and rand is a random number between 0 and 1. In the algorithm, N individuals are divided into two groups, namely leaders and followers. The leader is the first individual in the population, and its update formula is: In the formula, represents the position of the first salp in the d dimension, the leader, after t+1 iterations, is the location of the food in the d dimension, which is the location of the individual with the best fitness value, ub d lb d are the upper and lower boundaries of the d-dimensional space, is a random number between 0 and 1. is the convergence factor, and its expression is: Where t is the number of iterations and T is the maximum number of iterations; Except for the leader, the rest of the salps in the population are followers, which move forward in a chain, and the displacement conforms to Newton's laws of motion. The motion displacement formula is: Where a is acceleration, and the calculation formula is: a=(v fianl -v0) / T t , v0 is the initial velocity, T t is the time; in the iteration process T t =1, v0=0, and the final position formula of the follower is:

5. The method for predicting the working posture of intelligent engineering machinery according to claim 1, characterized in that: The particle swarm algorithm described above assumes that the search space is D-dimensional and has N particles, then the particle swarm is represented as: X = {x1, x2, ... x N }, where X represents the particle group, x i represents the i-th particle in the population, then the position of the i-th particle can be expressed as: i =(a1,a2,…,a D ) T , the corresponding velocity of the ith particle is: v i =(v1,v2,…,v D ) T , the local optimal solution P of the i-th particle best For: P i =(P1,P2,…,P D ) T , the global optimal solution G of the entire particle swarm best For: G i =(G1,G2,…,G D ) T ; During the iteration process, the particle updates its position and speed according to the current local optimal solution and the global optimal solution. The speed update formula is: v i (t+1)=ω·v i (t)+c1r1(P i (t)-a i (t))+c2r2(G i (t)-a i (t)), The position update formula is: a i (t+1)=a i (t)+v i (t+1), Where t represents the tth iteration of the particle, P i (t) represents the local optimum of the particle after the tth update, G i (t) represents the global optimum of the particle after the tth update, v i (t+1) represents the velocity of particle i after t+1 iterations, a i (t+1) represents the position of particle i after t+1 iterations. c1 and c2 are learning factors, which are individual learning factor and social learning factor respectively. r1 and r2 are random numbers between 0 and 1, which are used to represent the randomness of each particle in the search process of the particle swarm algorithm and increase the possibility of reaching the optimal solution. The inertia coefficient ω is used to balance the local search and the global search.

6. The method for predicting the working posture of intelligent engineering machinery according to claim 1, characterized in that: The method of increasing the search space and obtaining the improved salp algorithm through multiple cycles includes: firstly using the salp algorithm to find the optimal solution, It represents the difference between the food position after t+1 iterations and the food position after t iterations in the d dimension divided by the food position after t iterations; when the salp algorithm has λ(t+1)<1% for 20 consecutive times, the iteration change is very small, and the algorithm is randomly perturbed, that is, the salp algorithm fused with the particle swarm algorithm is used for optimization; then the search space of the salp algorithm is increased, and the upper limit of the d-dimensional space is set to In the formula, is the upper limit of the expanded d-dimensional space, ub d is the upper limit before the d-dimensional space is expanded, t is the number of iterations, and T is the maximum number of iterations; continue to search for the best solution, compare the optimal solution obtained by the search with the optimal solution obtained by the Salp Sea Squid algorithm to obtain a new optimal solution, and use N zhq As the cumulative number, when N zhq When ≥3 or t=T reaches the maximum number of iterations, the iteration stops and the final optimal solution is obtained, and the improved Salp Algorithm is obtained.

7. The method for predicting the working posture of intelligent engineering machinery according to claim 1, characterized in that: The extended Kalman filter algorithm optimizes the nonlinear system by using Taylor expansion and retaining only the first-order Taylor series expansion term, and then estimates the state information of the system through the state equation and observation equation of the system, uses the last optimal result to predict the current value, and uses the observed value to correct the current value to obtain the optimal result; the state equation and measurement equation of the system are: In the formula, the nonlinear functions f(·) and h(·) reflect the state vector and the measurement vector The mapping relationship between k-1 is the process noise, v k is the measurement noise, which is independent of each other and obeys N(0,Q k ) and N(0,R k )’s Gaussian white noise; The basic formula of the extended Kalman filter algorithm is: In the formula, represents the prior estimate of time k, and its corresponding prior error covariance matrix is represents the optimal estimate at time k-1, and its corresponding covariance matrix is μ k-1 is the system input at time k-1, Q k-1 is the covariance matrix of the system process noise at time k-1, A k For the equation The Jacobian matrix of ; In the formula, K k is the Kalman gain matrix, H k For the equation The Jacobian matrix, R k represents the covariance matrix of the measurement noise at time k, represents the optimal estimate at time k, and its corresponding covariance matrix is 8. The method for predicting the working posture of intelligent engineering machinery according to claim 1, characterized in that: The method of optimizing the system noise covariance matrix in the extended Kalman filter algorithm by using the improved salp algorithm to obtain an intelligent extended Kalman filter algorithm includes: optimizing the system noise covariance matrix by using the improved salp algorithm to obtain a modified covariance matrix as a new system noise covariance matrix in the extended Kalman filter, and using the extended Kalman filter to perform parameter estimation, and finally obtaining an optimized state estimation value; Furthermore, the use of the intelligent extended Kalman filter algorithm to predict the engineering operation posture includes: solving the angular velocity information and acceleration information measured by the inertial measurement unit to obtain the working posture of the engineering machinery, inputting the working posture of the engineering machinery into the intelligent extended Kalman filter algorithm, and predicting the working posture of the engineering machinery.

9. An intelligent engineering machinery operation posture prediction system, characterized in that: include: The operation posture information acquisition module is configured to: acquire the current operation posture information of the construction machinery; The operation posture information prediction module is configured as follows: based on the current operation posture information of the construction machinery, an intelligent extended Kalman filter algorithm is used to predict the next operation posture of the construction machinery, specifically: the salp algorithm is used for optimization, when the salp algorithm falls into a local optimal state, the particle swarm algorithm is integrated into the salp algorithm and the search space is increased, and the improved salp algorithm is obtained by multiple cycles, and the improved salp algorithm is used to optimize the system noise covariance matrix in the extended Kalman filter algorithm to obtain the intelligent extended Kalman filter algorithm, and the intelligent extended Kalman filter algorithm is used to predict the engineering operation posture; The particle swarm algorithm is integrated into the salp algorithm, including: introducing the speed update formula and position update formula of the particle swarm algorithm to improve the update formula of the leader and the follower in the salp algorithm; The update formula of the leader after the improvement of the Salp Algorithm is: In the formula, They represent the position of the first salp in the d dimension, i.e. the leader, after t and t+1 iterations, respectively. is the location of the food in the d dimension, which is the location of the individual with the best fitness value, ub d lb d are the upper and lower boundaries of the d-dimensional space, is a random number between 0 and 1. is the convergence factor, and its expression is: Where t is the number of iterations, T is the maximum number of iterations, a and b are the weight coefficients of the Salp Algorithm and the Particle Swarm Optimization Algorithm, respectively. And a+b=1; V1 d (t+1) represents the speed of the leader in the d dimension at the t+1th iteration, and its expression is: Where ω is the inertia weight, r1 and r2 are two random numbers, ε1 and ε2 are learning factors, is the individual optimal solution, is the optimal solution for the population; After the improvement of the Salp Algorithm, the update formula of the follower becomes: In the formula, denote the position of the i-th salp in d dimension after t and t+1 iterations, respectively. i d (t+1) represents the speed of the ith follower in the d-dimensional space at the t+1th iteration, ω is the inertia weight, r1 and r2 are two random numbers, ε1 and ε2 are learning factors, is the individual optimal solution, is the optimal solution for the population, T t is the time, which is 1 during the iteration.

Images