A multi-node path planning method for electric tractors

Abstract

The present invention provides a multi-node path planning method for electric tractors, belonging to the field of multi-node path planning for electric tractors. The method utilizes an improved whale algorithm to obtain operating parameters of an improved ant colony algorithm. The operating parameters are substituted into the improved ant colony algorithm to solve the multi-node optimal path. The result of solving the multi-node optimal path by the improved ant colony algorithm is then fed into the iteration of the improved whale algorithm. An evaluation function is provided in the improved whale algorithm, and the result of the improved ant colony algorithm is evaluated using the evaluation function of the improved whale algorithm. The present invention has the beneficial effects of fast convergence speed, strong convergence capability, short planned electric tractor operation paths, and low energy consumption.

Description

Technical Field

[0001] The invention relates to the field of multi-node path planning for electric tractors, and in particular to a multi-node path planning method for electric tractors. Background Art

[0002] Electric tractors have the advantages of low pollution, low noise, high efficiency and easy operation. In recent years, electric tractors have been mostly used in agricultural production. In order to reduce operating losses and improve endurance, multi-node path planning is used to improve operating efficiency. Multi-node path planning means that there are multiple nodes in a certain area, and starting from a certain node, the path after traversing all nodes is optimized according to the set operating rules.

[0003] Currently, multi-node path planning often uses intelligent algorithms, such as the ant colony algorithm (ACO) and the whale optimization algorithm (WOA). For example, in the article "An Improved Ant Colony Algorithm for the Traveling Salesman Problem" published in the journal Computer Applications, the algorithm was used to solve the traveling salesman problem. Because the ACO is based on the foraging behavior of ants, which is similar to path planning, it is a commonly used algorithm for multi-node path planning. When applying the ACO to multi-node path planning, it is necessary to set iteration rules for the parameters of multiple nodes. However, the ACO requires many parameters, and setting an excessively large number of iteration terminations can lead to multiple iterations with the same number of iterations in the later stages of the iteration, reducing solution efficiency. However, the termination number cannot be set too low, making it difficult to determine the optimal parameter combination. Therefore, the algorithm is prone to falling into local optimal solutions, which affects the solution of the optimal path.

[0004] Due to the problems of the ant colony algorithm, the whale optimization algorithm is often used to solve the optimal solution of multi-node path planning. Although the whale optimization algorithm has the advantages of a small number of setting parameters and strong convergence performance, it has the disadvantages of easily falling into local optimality and low convergence accuracy and slow speed. Although the improved whale optimization algorithm and its application in "Computer Engineering and Applications" and the hybrid strategy improved whale optimization algorithm in "Computer Application Research" have optimized the convergence ability of the whale optimization algorithm and enhanced the diversity of initial solutions by improving the whale algorithm, they have not solved the problems of low convergence accuracy and slow convergence speed. On the contrary, the complexity of the running logic has led to a decrease in the running efficiency of the algorithm.

[0005] Because it is impossible to take into account both local and global search capabilities, the path length is long and the energy consumption is high. Therefore, some electric tractor multi-node path planning methods currently use a combination of ant colony algorithm and basic particle swarm optimization algorithm (Partricle Swarm Optimization, PSO) to solve the problems of the ant colony algorithm. However, the problem of reduced efficiency due to the large number of iterations of the ant colony algorithm and the slow convergence speed of the whale optimization algorithm are not well solved, and there is also a certain dependence on terrain data. Summary of the Invention

[0006] This invention provides a multi-node path planning method for electric tractors with fast convergence speed, high convergence accuracy, and balanced local and global search capabilities in various terrains. The method is particularly suitable for multi-node path planning for electric tractors used in agricultural production. The technical problems addressed by this invention include the reduced efficiency of the ant colony algorithm due to excessive iterations, the slow convergence speed of the whale optimization algorithm, and the inability of the optimized algorithm to effectively balance global and local search in various terrains.

[0007] The technical solution adopted by the present invention is: a multi-node path planning method for an electric tractor, the method is to integrate the improved whale-ant colony algorithm, use the improved whale algorithm to obtain the operating parameters of the improved ant colony algorithm, substitute the operating parameters into the improved ant colony algorithm to solve the multi-node optimal path, put the result of the improved ant colony algorithm solving the multi-node optimal path into the iteration of the improved whale algorithm, set an evaluation function in the improved whale algorithm, and evaluate the result of the improved ant colony algorithm by the evaluation function of the improved whale algorithm. The working process is as follows:

[0008] The first step is to initialize the algorithm and import data;

[0009] The second step is to operate the imported data through the population initialization algorithm of the reverse learning strategy in the improved whale algorithm;

[0010] The third step is to input the calculation results into the improved ant colony algorithm as the operation parameters and initialize the improved ant colony algorithm;

[0011] The fourth step is to import node data and perform iterative calculations on path optimization;

[0012] Step 5: Determine whether the result of the iterative calculation of the path optimization has traversed all nodes. If the result of the iterative calculation of the path optimization has not traversed all nodes, return to step 4. If the result of the iterative calculation of the path optimization has traversed all nodes, update the pheromone.

[0013] The sixth step is to introduce the updated pheromone into the algorithm of the improved ant colony algorithm that introduces the iterative early termination strategy to determine whether the iterative early termination strategy is met. If the conditions of the iterative early termination strategy are met, the evaluation function is used to calculate and output the evaluation function value. If the conditions of the iterative early termination strategy are not met, it is determined whether the number of iterations of the algorithm that introduces the early termination strategy in the improved ant colony algorithm has reached the maximum value. If the number of iterations reaches the maximum value, the evaluation function is used to calculate and output the evaluation function value. If the number of iterations does not reach the maximum value, the current iteration result of the improved ant colony algorithm is used as a parameter and returned to the fourth step.

[0014] Step 7: Import the recorded optimization data into step 4 of the improved whale algorithm to record the evaluation function value and locate the individual position with the best evaluation function at the optimal position, and continue to improve the subsequent steps of the whale algorithm;

[0015] Step 8: Determine whether the number of iterations of the improved whale swarm algorithm in step 7 has reached the maximum number of iterations. If the number of iterations reaches the maximum number of iterations, output the optimal parameters and fitness value and terminate the operation. The fitness value is the evaluation function value. If the number of iterations does not reach the maximum number of iterations, return to step 3 and input the result parameters of the improved whale swarm algorithm in step 7 into step 3 as the operation parameters.

[0016] Furthermore, the data imported in the first step include improved ant colony algorithm variables and improved whale algorithm variables. The improved ant colony algorithm variables include pheromone importance factor, heuristic function importance factor, pheromone concentration volatility factor and the number of ants in the ant colony. The improved whale algorithm variables include the number of whale populations and the maximum number of iterations.

[0017] Furthermore, the algorithm flow of introducing the iteration termination strategy in the improved ant colony algorithm is as follows:

[0018] The first step is to import the iteration data;

[0019] The second step is to determine whether the data of the last two iterations are the same. If they are not the same, the number of times the iterative solutions are the same is set to zero and the iteration is continued. If the data of the last two iterations are the same, the number of times the iterative solutions are the same is increased by one.

[0020] The third step is to determine whether the number of consecutive identical iterations is greater than or equal to the set value. If it is less than the set value, continue the iteration. If it is greater than or equal to the set value, proceed to the next step.

[0021] The fourth step is to determine whether the iterative solution is less than or equal to the iterative solution set value. If the iterative solution is greater than the iterative solution set value, the iteration continues. If the iterative solution is less than or equal to the set value, the iteration is terminated and the solution is output.

[0022] Further, the steps to improve the whale group algorithm are:

[0023] Step 1: Determine the variable types, population number, and maximum number of iterations of the improved ant colony algorithm;

[0024] Step 2: Randomly initialize the population within the variable value range;

[0025] Step 3: Establish the initial population by introducing the reverse learning strategy.

[0026] Step 4: Record the individual position with the best evaluation function value as the optimal position;

[0027] Step 5: Complete the individual position update based on the encirclement and predation search mechanism, spiral ascending mechanism or random search mechanism, introduce the random variable p = rand (0, 1), when p < 0.5 and |A| < 1, the individual updates its position according to the encirclement and predation search mechanism, when p < 0.5 and |A| ≥ 1, the individual updates its position according to the random search mechanism, when p ≥ 0.5, the individual updates its position according to the spiral ascending mechanism.

[0028] Furthermore, the encirclement and predation search mechanism formula is as follows:

[0029] D=|CX * (t)-X(t)|

[0030] X(t+1)=wX * (t)-AD

[0031]

[0032] A=2ar-a=a(2r-1)

[0033] C=2r

[0034]

[0035] Where t is the current iteration number, X is the current position, and X * is the optimal position, A and C are coefficient vectors, r is a random variable with a range of [0, 1], a is the nonlinear convergence factor, T max is the maximum number of iterations, w is the adaptive inertia weight factor, D is the difference between the optimal position and the current position,

[0036] The spiral rising mechanism formula is:

[0037] X(t+1)=De bl cos(2πl)+wX * (t)

[0038] Where b is a constant, l is a random variable with a range of [0, 1], t is the current iteration number, X is the current position, and X * For the best position

[0039] The random search mechanism formula is:

[0040] D * =|CX rand (t)-X(t)|

[0041] X(t+1)=X rand (t)-AD *

[0042] Where, X rand (t) is a randomly selected position in the population.

[0043] Furthermore, the population initialization algorithm of the reverse learning strategy establishes the method of population initialization as follows:

[0044] The first step is to establish a random initial population and analyze the individuals in the random initial population one by one;

[0045] In the second step, a random number p is introduced. If p≥0.3, reverse learning is performed. If p<0.3, no reverse learning is performed and the population is directly determined as the final initial population.

[0046] The formula for reverse learning is:

[0047]

[0048] Where k is the subscript number, the minimum value range is 1, and the maximum value range is the number of variables in the improved ant colony algorithm. k and U k is the population space boundary, X r is the position of individual r in d-dimensional space, For X r The corresponding reverse individual,

[0049] Ant number boundary: L1 = 20 ~ 30, U1 = 80 ~ 90,

[0050] Pheromone volatility factor boundary: L2 = 0.2 ~ 0.25, U2 = 0.75 ~ 0.8,

[0051] Heuristic function importance factor boundary: L3 = 1 ~ 2, U3 = 5 ~ 6,

[0052] Pheromone importance factor boundary: L4 = 0.3 ~ 0.5, U4 = 2 ~ 2.5,

[0053] Calculate X r and The evaluation function value of the final initial population is retained as the population with the highest evaluation function value.

[0054] Furthermore, the state transfer function formula of the improved ant colony algorithm for path selection is:

[0055]

[0056] Updated pheromone formula: Ta ij (t ACA )=(1-rh)Ta ij (t ACA -1)+ΔTa ij (t ACA )

[0057] The parameter in the formula is the pheromone update parameter, and the update is the iterative update

[0058]

[0059] Where: t ACA is the current iteration number, is the state transfer function, Ta ij (t ACA ) is the pheromone concentration, α is the pheromone importance factor, Et ij (t ACA ) is the heuristic function, β is the importance factor of the heuristic function, rh is the pheromone concentration volatility factor, ΔTa ij (t ACA ) is the new amount of pheromone on the path, which is related to whether the path is passed. Dis ij is the Euclidean distance of the path, i and j are nodes, and allow represents the nodes that have not been passed through.

[0060] Furthermore, the evaluation function reflects the optimality of the objective function, rapid convergence and algorithm stability of the improved ant colony algorithm for solving multi-node path planning. The formula of the evaluation function is:

[0061] y=k1f1+k2f2+k3f3+k4f4

[0062] f1 = length - length_min

[0063] f2=energy-energy_min

[0064] f3=Σlength

[0065] f4=σ(length)

[0066] Where: length is the path length value solved by the improved ant colony algorithm, length_min is the empirical optimal value of the path length, energy is the path energy consumption value solved by the improved ant colony algorithm, energy_min is the empirical optimal value of the path energy consumption, f1 is the difference between the path length value solved by the improved ant colony algorithm and the empirical optimal value, f2 is the difference between the energy consumption value solved by the improved ant colony algorithm and the empirical optimal value, f1 and f2 represent the optimality of the algorithm in solving the objective function, f3 is the number of iterations of the improved ant colony algorithm, which represents the rapid convergence of the algorithm, f4 is the standard deviation of the iterative data of the improved ant colony algorithm, which represents the stability of the iterative data of the algorithm, and k1, k2, k3, and k4 are weight coefficients.

[0067] The above structure has the following beneficial effects:

[0068] 1. Since the whale algorithm and the ant colony algorithm are optimized at the same time, and the improved ant colony algorithm is put into the iteration of the improved whale algorithm, the early termination strategy of the improved ant colony algorithm is used to overcome the problem of too many iterations, thereby improving the efficiency of the algorithm. In addition, the population initialization algorithm of the reverse learning strategy in the improved whale algorithm is used to improve the quality of the initial population, thereby improving the convergence speed and accuracy of the algorithm.

[0069] 2. The introduction of nonlinear convergence factors prevents the algorithm from switching from global search to local search too early, thus overcoming the problem of the algorithm falling into the local optimal solution. The introduction of adaptive inertia weight factors overcomes the problem of the whale optimization algorithm easily falling into the local optimal solution, thereby enhancing the global search capability of the algorithm in the early iteration. The smaller adaptive weight factor in the later iteration reduces the individual's dependence on the current optimal solution, thus avoiding the algorithm from falling into the local optimal solution.

[0070] 3. The fusion-improved whale-ant colony algorithm introduces an early termination strategy for iteration, population initialization of the reverse learning strategy, a nonlinear convergence factor, and an adaptive weight factor. It overcomes the problems of excessive iterations, local-global imbalance, low convergence accuracy, and low efficiency, improves the convergence speed and accuracy, and effectively reduces energy loss and path distance in the planned path, balancing the local and global search capabilities under different road conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 This is the flow chart of the fusion-improved whale-ant colony algorithm of the present invention;

[0072] Figure 2 It is a flow chart of the algorithm for improving the iterative early termination strategy in the ant colony algorithm of the present invention;

[0073] Figure 3 is a graph of the nonlinear convergence factor of the present invention;

[0074] Figure 4 It is a Sphere function convergence curve diagram of the improved whale algorithm performance test of the present invention;

[0075] Figure 5 It is a convergence curve diagram of the Quartic function of the improved whale algorithm performance test of the present invention;

[0076] Figure 6 It is a Schwefel 2.26 function convergence curve diagram of the improved whale algorithm performance test of the present invention;

[0077] Figure 7 It is a Rastrigin function convergence curve diagram of the improved whale algorithm performance test of the present invention;

[0078] Figure 8 is a force analysis diagram of the electric tractor motion model of the present invention;

[0079] Figure 9 It is a position diagram of 26 nodes in a flat space in the flat environment simulation of the present invention;

[0080] Figure 10 It is a convergence curve diagram of the evaluation function in the flat environment simulation of the present invention;

[0081] Figure 11 It is a path planning diagram of the integrated particle swarm and ant colony algorithm in the flat environment simulation of the present invention;

[0082] Figure 12 It is a path planning diagram of the integrated improved whale-ant colony algorithm in the flat environment simulation of the present invention;

[0083] Figure 13 is a path length convergence curve diagram of the flat environment simulation of the present invention;

[0084] Figure 14 It is a spatial position diagram of 31 nodes in the rugged environment simulation of the present invention;

[0085] Figure 15 is a graph of convergence of the evaluation function of the rugged environment simulation of the present invention;

[0086] Figure 16 It is a path planning diagram of the present invention integrating particle swarm and ant colony algorithm in rugged environment simulation;

[0087] Figure 17 It is a path planning diagram of the present invention integrating the improved whale-ant colony algorithm in rugged environment simulation;

[0088] Figure 18 It is a convergence curve diagram of the evaluation function of the rugged environment simulation of the present invention. DETAILED DESCRIPTION

[0089] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0090] like Figure 1-2 As shown:

[0091] The improved ant colony algorithm introduces an early termination strategy for iteration. When the set number of iteration terminations is larger than the problem being solved, multiple iteration values will be the same in the later stages of the iteration, which reduces the algorithm's solution efficiency. Therefore, the early termination strategy introduced by the improved ant colony algorithm ensures that the iteration is terminated when the number of consecutive iterative solutions reaches the set value and the iterative solution is less than the set value. Otherwise, the iteration is terminated when the maximum number of iterations is met. This ensures the early termination of the iteration, avoids invalid iterations, and speeds up the algorithm efficiency.

[0092] The population initialization algorithm of the reverse learning strategy used in the Improved Whale Optimization Algorithm (IWOA) improves the population diversity while improving the population quality. The high-quality initial population helps the algorithm converge quickly and avoids the convergence speed being reduced due to the initial population quality.

[0093] like Figure 3 As shown in the figure, the nonlinear convergence factor introduced in the improved whale algorithm better balances the global search and local search capabilities of the algorithm, so that the algorithm switches to local search after fully performing global search.

[0094] An adaptive inertia weight factor is introduced in the improved whale algorithm, so that individuals in the population consider the influence of their current position when moving towards the local optimal solution. Because in the whale optimization algorithm, the inertia weight factor can be regarded as a constant 1. A larger inertia weight factor in the early iteration helps the global search of the algorithm. In the later iteration, due to the excessively large weight factor, individuals will pay too much attention to the current optimal solution position, and thus easily fall into the local optimal solution. Therefore, the adaptive weight factor w is introduced, and the value range of w is [0.5,0.9]. The value of w in the early iteration is close to 0.9, which enhances the global search ability of the algorithm. The value of w in the later iteration is close to 0.5, which reduces the dependence on the current optimal solution and avoids falling into the local optimal solution.

[0095] Improved performance test of the whale algorithm:

[0096] Four functions are selected for testing, as shown in Table 1. The population size is set to 40 and the maximum number of iterations is set to 100. The convergence effects of the improved whale algorithm, whale optimization algorithm and basic particle swarm algorithm are compared.

[0097] Table 1 Test functions

[0098]

[0099] Sphere and Quartic are unimodal functions used to test the local search ability of the algorithm; Schwefel 2.26 and Rastrigin are multimodal functions used to test the global search ability of the algorithm;

[0100] like Figure 4-7 As shown in the convergence curve diagram of the improved whale algorithm, whale optimization algorithm and basic particle swarm algorithm applied to the test function, the improved whale algorithm has converged to the optimal value in the 10th to 20th iterations, and its convergence speed is better than that of the whale optimization algorithm and basic particle swarm algorithm.

[0101] The improved whale algorithm, whale optimization algorithm and basic particle swarm algorithm were run 30 times on the four test functions respectively, and the optimal value, worst value and standard deviation were recorded. The results are shown in Table 2 Test function result analysis.

[0102] Table 2 Test function result analysis

[0103]

[0104]

[0105] According to the data in the table, the optimal and worst values of the function obtained by the improved whale algorithm are better than those of the whale optimization algorithm and the basic particle swarm algorithm. At the same time, the standard deviation of the improved whale algorithm is lower and the performance is more stable. Therefore, the performance of the improved whale algorithm in both unimodal and multimodal test functions is better than that of the whale optimization algorithm and the basic particle swarm algorithm. It is concluded that the local search and global search performance of the improved whale algorithm are better, verifying the optimization effect.

[0106] Multi-node path planning simulation for electric tractors:

[0107] In agricultural production, a node is a gathering point for nearby harvested crops. An electric tractor needs to traverse all nodes in the area to collect all harvested crops. The simulation object is a certain cultivated land. Node data and the operating parameters of the electric tractor are introduced to simulate and verify the optimization performance of the fusion-improved whale-ant colony algorithm.

[0108] Excluding the influence of irrelevant factors on the experiment, the following assumptions are made:

[0109] (1) The path between nodes is a straight line segment;

[0110] (2) When calculating the path length and energy loss, the steering action of the electric tractor at the node is ignored;

[0111] (3) Ignore the air resistance of the electric tractor;

[0112] (4) Focus on the path length and energy loss of the electric tractor traveling between nodes.

[0113] Algorithm simulation environment: operating system Windows 10 (64-bit), processor Core(TM) i7-8550U, CPU 1.80 GHz, running memory 16 GB, simulation platform Matlab R2017a.

[0114] The operating parameters of the electric tractor are: the friction coefficient of the road surface is 0.07, the gravity of the electric tractor is 10700N, and the transmission efficiency is 0.85.

[0115] When an electric tractor travels between different operating points, it usually travels at a low and constant speed. The energy loss during the operation of the electric tractor is mainly the energy consumed during the travel. Figure 8 shown.

[0116] Electric tractor sports model:

[0117]

[0118]

[0119] Where s i,j is the horizontal distance between nodes, s i,j is the height difference between nodes, θ i,j is the inclination angle between nodes, γ i,j is the path length between nodes,

[0120] When the electric tractor is traveling at a constant speed, the required driving force F of the electric tractor motor is q The formula is

[0121] F q =μGcos(θ i,j )+Gsin(θ i,j )

[0122] Where: μ is the friction coefficient of the road surface, G is the gravity of the electric tractor, and G is assumed to be a constant without considering the change of the load weight.

[0123] The energy consumed by the motor during operation is expressed as follows:

[0124] Q i,j =q i,j

[0125] q i,j = i,j t

[0126] Where Q i,jis the energy consumed between the two nodes when the motor is running, U is the input voltage of the motor, q i,j is the power consumed by the motor between the two nodes, I i,j is the motor current, t is the starting time,

[0127] The motor driving force F output by the electric tractor during travel q

[0128]

[0129] Where, T x is the output torque of the planet carrier, i g is the transmission speed ratio of the electric tractor, i0 is the differential speed ratio of the electric tractor, η is the transmission efficiency, R is the wheel radius of the electric tractor, T t is the output torque of the motor,

[0130]

[0131] Where n is the motor speed, P is the motor output power,

[0132] The relationship between the travel speed of an electric tractor and the motor speed is:

[0133]

[0134] The formula of the integrated electric tractor motion model is used to obtain the energy consumed by the electric tractor running between nodes i and j:

[0135]

[0136] where γ i,j As the evaluation function, the improved ant colony algorithm solves the path length value, Q i,j The energy consumption value of the path is solved by the improved ant colony algorithm as the evaluation function.

[0137] Example 1: Flat environment simulation,

[0138] like Figure 9 As shown in the figure, the cultivated land near 87.4 east longitude and 44.3 north latitude is selected as the simulation object, and 26 nodes are selected in the flat space. The evaluation function is uniformly adopted.

[0139] y=k1f1+2f2+3f3+4f4

[0140] The algorithm parameters are set as follows: the number of variables in the improved ant colony algorithm is 4, the number of populations is 30, the maximum number of iterations is 50, k1=0.1, k2=0.00005, k3=0.1, k4=0.001.

[0141] In order to better analyze the fusion improved whale-ant colony algorithm, the fusion improved whale-ant colony algorithm and the fusion particle swarm-ant colony algorithm are used to solve the problem of multi-node path planning in a flat environment.

[0142] The convergence curves of the evaluation functions of the fusion improved whale-ant colony algorithm and the fusion particle swarm-ant colony algorithm are as follows: Figure 10 As shown in Table 3, the optimized ant colony algorithm parameter values are shown in Table 3.

[0143] Table 3 Parameter values of improved ant colony algorithm

[0144] algorithm m rh β α PSO-ACO 80 0.80 4.39 0.50 IWOA-ACO 80 0.80 5.00 0.50

[0145] Depend on Figure 8 It can be seen that the convergence value of the evaluation function of the fusion improved whale-ant colony algorithm is better than that of the fusion particle swarm-ant colony algorithm;

[0146] like Figure 11-12 As shown in the figure, when the starting nodes are the same, the arrows represent the direction of travel. The path length of the fusion improved whale-ant colony algorithm is 8750m, and the energy consumption is 2141837J. The path length of the fusion particle swarm-ant colony algorithm is 8750m, and the energy consumption is 2141837J.

[0147] like Figure 13 As shown in the figure, the path length and convergence speed of the fused improved whale-ant colony algorithm (IWOA-ACO) are better than those of the fused particle swarm-ant colony algorithm (PSO-ACO). At the same time, the final optimization result is optimal and the number of iterations is less than one-third of the maximum number of iterations, which ensures the convergence speed while improving the convergence accuracy.

[0148] The results of the fusion-improved whale-ant colony algorithm and the fusion-particle swarm-ant colony algorithm are the same under flat environment. However, the fusion-improved whale-ant colony algorithm under flat environment has the advantages of fast convergence speed and strong convergence ability of the fusion-particle swarm-ant colony algorithm, which helps to improve the operating efficiency of electric tractors.

[0149] Example 2: Rough Environment Simulation

[0150] In order to verify the adaptability of the fusion-improved whale-ant colony algorithm in diverse farmland environments, 31 node data in a rugged environment are introduced. The spatial positions of the 31 nodes are as follows: Figure 14 shown.

[0151] The convergence curves of the evaluation functions corresponding to the fusion improved whale-ant colony algorithm and the fusion particle swarm-ant colony algorithm are as follows: Figure 15 As shown in Table 4, the optimized ACO parameter values are as follows:

[0152] Table 4 Improved ant colony parameter values

[0153] parameter m rh β α PSO-ACO 47 0.2 4.69 0.55 IWOA-ACO 72 0.78 2 0.83

[0154] like Figure 4 As shown in the figure, the convergence value and convergence speed of the evaluation function of the fusion improved whale-ant colony algorithm are better than those of the fusion particle swarm-ant colony algorithm. The fusion improved whale-ant colony algorithm reaches the optimal value in about 5 iterations, with fast convergence speed and good convergence performance. It has obvious convergence performance improvement for the multi-node path planning problem on rugged paths.

[0155] The ant colony algorithm plans the path of the electric tractor at 31 nodes according to the parameters in Table 4. The optimized path is as follows: Figure 16-17 As shown in the figure, with the same starting point, the arrow indicates the forward direction. The path length of the fusion improved whale-ant colony algorithm is 19057m, the energy consumption is 4070478J, and the path length of the fusion particle swarm-ant colony algorithm is 19420m, the energy consumption is 4246522J.

[0156] The fusion improved whale-ant colony algorithm reduces the path length by 1.91% and the energy consumption by 4.32% compared with the fusion particle swarm-ant colony algorithm.

[0157] The final results show that the fusion-improved whale-ant colony algorithm has the shortest path length and the least energy consumption. It reduces the running length and reduces the energy loss in solving the multi-node path planning problem of electric tractors on rough roads, and is significantly better than the other two algorithms.

[0158] Analyzing the reasons, the number of individuals m=72 in the fusion-improved whale-ant colony algorithm enables the algorithm to converge quickly. At the same time, the pheromone concentration volatility factor rh=0.78 makes the algorithm have strong global convergence. The numerical difference between the pheromone importance factor α=0.83 and the heuristic function importance factor β=2 is small, which allows the iterative process to fully consider the pheromone concentration and heuristic function. In summary, the fusion-improved whale-ant colony algorithm can well balance global search and local search.

[0159] like Figure 18 As shown in Figure 3, according to the convergence curve of the evaluation function, it can be seen that the convergence speed block and path length of the fusion improved whale-ant colony algorithm are shorter, and the convergence performance is significantly better than that of the fusion particle swarm-ant colony algorithm.

[0160] Combining Example 1 and Example 2, it can be seen that the fusion-improved whale-ant colony algorithm can balance global and local searches while ensuring rapid convergence and convergence accuracy of the algorithm. Compared with the existing algorithm, the fusion-improved whale-ant colony algorithm can ensure the shortest path length and the lowest energy consumption regardless of whether it is on flat or rugged roads, effectively improving operating efficiency while reducing energy consumption, and achieving energy-saving and efficient effects in the use of electric tractors in agricultural work.

[0161] The directional terms mentioned in the present invention, such as "up", "down", "left", "right", etc., are only for better and clearer explanation and understanding of the present invention, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation. Therefore, they should not be understood as limiting the present invention.

[0162] The above description is based on the preferred embodiments of the present invention, but it should not be understood as limiting the claims. The present invention is not limited to the above embodiments, and its specific structure is allowed to be varied. All variations made within the scope of protection of the independent claims of the present invention are within the scope of protection of the present invention.

Claims

1. A multi-node path planning method for an electric tractor, characterized by: The method is to integrate the improved whale-ant colony algorithm, use the improved whale algorithm to obtain the operating parameters of the improved ant colony algorithm, substitute the operating parameters into the improved ant colony algorithm to solve the multi-node optimal path, put the result of the improved ant colony algorithm solving the multi-node optimal path into the iteration of the improved whale algorithm, set an evaluation function in the improved whale algorithm, and evaluate the results of the improved ant colony algorithm through the evaluation function of the improved whale algorithm. The workflow is as follows: The first step is to initialize the algorithm and import data; The second step is to operate the imported data through the population initialization algorithm of the reverse learning strategy in the improved whale algorithm; The third step is to input the calculation results into the improved ant colony algorithm as the operation parameters and initialize the improved ant colony algorithm; The fourth step is to import node data and perform iterative calculations on path optimization; Step 5: Determine whether the result of the iterative calculation of the path optimization has traversed all nodes. If the result of the iterative calculation of the path optimization has not traversed all nodes, return to step 4. If the result of the iterative calculation of the path optimization has traversed all nodes, update the pheromone. The sixth step is to introduce the updated pheromone into the algorithm of the improved ant colony algorithm that introduces the iterative early termination strategy to determine whether the iterative early termination strategy is met. If the conditions of the iterative early termination strategy are met, the evaluation function is used to calculate and output the evaluation function value. If the conditions of the iterative early termination strategy are not met, it is determined whether the number of iterations of the algorithm that introduces the early termination strategy in the improved ant colony algorithm has reached the maximum value. If the number of iterations reaches the maximum value, the evaluation function is used to calculate and output the evaluation function value. If the number of iterations does not reach the maximum value, the current iteration result of the improved ant colony algorithm is used as a parameter and returned to the fourth step. Step 7: Import the recorded optimization data into step 4 of the improved whale algorithm to record the evaluation function value and locate the individual position with the best evaluation function value at the optimal position, and continue to improve the subsequent steps of the whale algorithm; The eighth step is to determine whether the number of iterations of the improved whale swarm algorithm in the seventh step has reached the maximum number of iterations. If the number of iterations has reached the maximum number of iterations, the optimal parameters and fitness value are output and the operation is terminated. If the number of iterations has not reached the maximum number of iterations, the operation returns to the third step and the result parameters of the improved whale swarm algorithm in the seventh step are input into the third step as the running parameters.

2. The electric tractor multi-node path planning method according to claim 1, characterized in that: The data imported in the first step include improved ant colony algorithm variables and improved whale algorithm variables. The improved ant colony algorithm variables include pheromone importance factor, heuristic function importance factor, pheromone concentration volatility factor and the number of ants in the ant colony. The improved whale algorithm variables include the number of whale populations and the maximum number of iterations.

3. The multi-node path planning method for an electric tractor according to claim 1, characterized in that: The algorithm flow of introducing the iterative termination strategy in the improved ant colony algorithm is as follows: The first step is to import the iteration data; The second step is to determine whether the data of the last two iterations are the same. If they are not the same, the number of times the iterative solutions are the same is set to zero and the iteration is continued. If the data of the last two iterations are the same, the number of times the iterative solutions are the same is increased by one. The third step is to determine whether the number of consecutive identical iterations is greater than or equal to the set value. If it is less than the set value, continue the iteration. If it is greater than or equal to the set value, proceed to the next step. The fourth step is to determine whether the iterative solution is less than or equal to the iterative solution set value. If the iterative solution is greater than the iterative solution set value, the iteration continues. If the iterative solution is less than or equal to the set value, the iteration is terminated and the solution is output.

4. The multi-node path planning method for an electric tractor according to claim 1, characterized in that: The steps of improving the whale group algorithm are as follows: Step 1: Determine the variable types, population number, and maximum number of iterations of the improved ant colony algorithm; Step 2: Randomly initialize the population within the variable value range; Step 3: Establish the initial population by introducing the reverse learning strategy. Step 4: Record the individual position with the best evaluation function value as the optimal position; Step 5: Complete the position update of the individual based on the encirclement and predation search mechanism, spiral ascending mechanism or random search mechanism, introduce the random variable p = rand(0,1), when p < 0.5 and |A| < 1, the individual updates its position according to the encirclement and predation search mechanism, when p < 0.5 and |A| ≥ 1, the individual updates its position according to the random search mechanism, and when p ≥ 0.5, the individual updates its position according to the spiral ascending mechanism.

5. The multi-node path planning method for an electric tractor according to claim 4, characterized in that: The encirclement and predation search mechanism formula is as follows: D=|CX * (t)-X(t)| X(t+1)=wX * (t)-AD A=2ar-a=a(2r-1) C＝2r Where t is the current iteration number, X is the current position, and X * is the optimal position, A and C are coefficient vectors, r is a random variable with a range of [0,1], a is the nonlinear convergence factor, T max is the maximum number of iterations, w is the adaptive inertia weight factor, D is the difference between the optimal position and the current position, The spiral rising mechanism formula is: X(t+1)=De bl cos(2πl)+wX * (t) Where b is a constant, l is a random variable with a range of [0,1], t is the current iteration number, X is the current position, and X * For the best position The random search mechanism formula is: D * =|CX rand (t)-X(t)| X(t+1)=X rand (t)-AD * Where, X rand (t) is a randomly selected position in the population.

6. The multi-node path planning method for an electric tractor according to claim 4, characterized in that: The population initialization algorithm of the reverse learning strategy establishes the method of population initialization as follows: The first step is to establish a random initial population and analyze the individuals in the random initial population one by one; In the second step, a random number p is introduced. If p≥0.3, reverse learning is performed. If p<0.3, no reverse learning is performed and the population is directly determined as the final initial population. The formula for reverse learning is: Where k is the subscript number, the minimum value range is 1, and the maximum value range is the number of variables in the improved ant colony algorithm. k and U k is the population space boundary, X r is the position of individual r in d-dimensional space, For X r The corresponding reverse individual, Calculate X r and The evaluation function value of the final initial population is retained as the population with the highest evaluation function value.

7. The multi-node path planning method for an electric tractor according to claim 1, characterized in that: The state transfer function formula of the improved ant colony algorithm for selecting a path is: Are ij (t ACA )=(1-rh)Yes ij (t ACA −1)+△Ta ij (t ACA ) Where: t ACA is the current iteration number, is the state transfer function, Ta ij (t ACA ) is the pheromone concentration, α is the pheromone importance factor, Et ij (t ACA ) is the heuristic function, β is the importance factor of the heuristic function, rh is the pheromone concentration volatility factor, △Ta ij (t ACA ) is the new amount of pheromone on the path, which is related to whether the path is passed. Dis ij is the Euclidean distance of the path, i and j are nodes, and allow represents the nodes that have not been passed through.

8. The electric tractor multi-node path planning method according to claim 1, characterized in that: The formula of the evaluation function is: y=k1f1+k2f2+k3f3+k4f4 f1 = length - length_min f2=energy-energy_min f3=∑length f4=δ(length) Where: length is the path length value solved by the improved ant colony algorithm, length_min is the empirical optimal value of the path length, energy is the path energy consumption value solved by the improved ant colony algorithm, energy_min is the empirical optimal value of the path energy consumption, f1 is the difference between the path length value solved by the improved ant colony algorithm and the empirical optimal value, f2 is the difference between the energy consumption value solved by the improved ant colony algorithm and the empirical optimal value, f1 and f2 represent the optimality of the algorithm in solving the objective function, f3 is the number of iterations of the improved ant colony algorithm, which represents the rapid convergence of the algorithm, f4 is the standard deviation of the iterative data of the improved ant colony algorithm, which represents the stability of the iterative data of the algorithm, and k1, k2, k3, and k4 are weight coefficients.

Images