Maize harvesting machine in-field operation path planning method for soil mechanical compaction and reduction

Abstract

The invention discloses an in-field operation path planning method of a corn harvester for soil mechanical compaction and reduction, belongs to the field of agricultural machinery dispatching, and aims to solve the problem that soil compaction is intensified due to the fact that a path is not planned for in-field mechanical operation. The method comprises the following steps: acquiring furrow and boundary point information of a farmland to be planned; arranging a sensor; judging whether the to-be-planned farmland is a regular field, and if so, constructing an in-field path planning model of the regular field; solving the in-field path planning model of the regular field by using a grey wolf optimization algorithm, and outputting an optimal path of the regular field operation; if the field parcel is not the regular field parcel, judging whether the turning radius is smaller than the maximum turning radius allowed by the field parcel or not, if yes, executing an I-type turning method, if not, executing an II-type turning method, obtaining optimal turning path parameters under the two turning methods by using a cuckoo search algorithm, and performing path planning on the irregular field parcel by using an immune algorithm; and outputting the optimal path of the irregular field operation.

Description

Technical Field

[0001] The invention relates to field operation path planning and belongs to the field of agricultural machinery scheduling. Background Art

[0002] With the continuous advancement of agricultural modernization and the rapid development of artificial intelligence and smart agricultural technologies, intelligent operation of agricultural machinery has become an important direction for improving agricultural production efficiency and sustainable development. Black soil resources are extremely precious. They are highly fertile and loose, but also very fragile. They are easily compacted by mechanical operations, which affects soil permeability, water retention and crop growth environment, and ultimately reduces food production and sustainable soil utilization.

[0003] In the corn production process, traditional mechanized harvesting is often accompanied by high-intensity soil compaction. Due to the heavy weight of the harvester, the repeated rolling of the soil while driving in the field not only aggravates the soil compaction, but also causes soil structure damage under long-term use. To address this problem, agricultural production is gradually transforming towards intelligent operation. By optimizing the path planning of agricultural machinery in the field, the impact of soil compaction can be reduced, achieving the goal of high efficiency and low loss operation. Summary of the invention

[0004] In view of the problem that unplanned paths for mechanized field operations will aggravate soil compaction, the present invention provides a method for planning the field operation path of a corn harvester for reducing mechanical soil compaction.

[0005] The method for planning a corn harvester's field operation path for reducing soil mechanical compaction according to the present invention comprises the following steps:

[0006] Step 1: Obtain GNSS-based in-field ridge and ditch information and boundary point information of the farmland to be planned;

[0007] Step 2: Arrange sensors;

[0008] Step 3: Determine whether the farmland to be planned is a regular field. If it is a regular field, execute step 4; if it is not a regular field, execute step 6;

[0009] Step 4: Construct a path planning model for regular fields;

[0010] Step 5: Use the Gray Wolf Optimization Algorithm to solve the path planning model in the regular field and output the optimal path for the regular field operation;

[0011] Step 6: Determine whether the turning radius is less than the maximum turning radius allowed by the plot. If so, execute the Type I turning method, and then execute Step 7; if not, execute the Type II turning method, and then execute Step 8;

[0012] Step 7: Use the cuckoo search algorithm to obtain the optimal turning path parameters under the I-type turning method, and then execute step 9;

[0013] Step 8: Using the cuckoo search algorithm to obtain the optimal turning path parameters under the type II turning method;

[0014] Step 9: Use the immune algorithm to plan the path for irregular fields and output the optimal path for irregular field operations.

[0015] Preferably, the construction of the path planning model for the regular field in step 4 is as follows:

[0016] Objective Function In the formula, α, β are the weight coefficients of the objective function;

[0017] y i Indicates whether the variable is harvested.

[0018] z i,j Indicates whether the ridge pair (i, j) is compacted variable,

[0019] Decision variables: represents the path of the kth corn harvester, where p k,t represents the ridge pair of the k-th corn harvester's t-th corn harvester path, t∈{1,2,...,T k}, T k represents the maximum number of operation steps of each corn harvester; each p k,t ={i,j} represents the two ridges compacted in the current operation;

[0020] Constraints:

[0021] Condition 1: Harvest coverage constraints: x k,t,i Indicates whether the k-th corn harvester harvests the i-th ridge at the t-th harvest. Each ridge must be harvested once, and N is the total number of ridges in the field;

[0022] Condition 2: Compaction minimization constraint: x k,t,j Indicates whether the k-th corn harvester compacts the sensor of the j-th ridge during the t-th harvest, limiting the number of compactions of the sensor of each ridge to at most once, and S is the sensor layout plan;

[0023] Fixed compaction ridge constraint: Among them, K is the harvester set, and the corn harvester only compacts two ridges each time it travels;

[0024] A ridge constraint that spans but does not compact: Corn harvester can cross R on the path k Make ridges, but compact only 2 of them.

[0025] Preferably, step 5 is based on the path P of the kth corn harvester. k As the initial population individual of the gray wolf optimization algorithm, it iteratively outputs the optimal path for regular field operations.

[0026] Preferably, the turning radius R in step 6 is less than the maximum turning radius R allowed by the plot. max When the I-type turning method is used, the turning path parameters h′, r′, d′, and the path length Wherein, h' is the projection of the distance between the starting and ending points of the planned path in the I-type turning method in the planting direction of the field ridge, r' is the radius of the virtual turning path in the I-type turning method, d' is the distance between the field ridges inside the field in the I-type turning method, and e is the center distance between the front and rear tire spindles of the corn harvester;

[0027] The turning radius R in step 6 is greater than the maximum turning radius R allowed by the plot. max When the type II turning method is used, the turning path parameters are h″, r″, d″, and the path length Occupied land area

[0028] Where h″ is the projection of the distance between the starting and ending points of the planned path in the type II turning method in the planting direction of the ridges, r″ is the radius of the virtual turning path in the type II turning method, and d″ is the distance between the ridges inside the field in the type II turning method.

[0029] Preferably, in step nine, an immune algorithm is used to plan paths for irregular fields, with the operation sequence being the path of a corn harvester, and the optimal path for the irregular field operation is outputted through iteration.

[0030] Beneficial effects of the present invention: The present invention proposes a method for planning the operation path of a corn harvester in the field with the purpose of minimizing the compaction degree of the soil. By arranging sensors in the field to measure the compaction degree of each ridge in the field, the driving path of the corn harvester on each ridge on the soil is optimized. At the same time, based on multi-agent optimization algorithms (such as artificial fish swarm algorithm, gray wolf optimization algorithm, etc.), efficient solution of path planning is achieved to ensure that the impact of soil compaction is minimized under the premise of high field coverage, and the operation goal of high efficiency and low loss is achieved. This method can not only effectively protect soil resources, but also provide technical reference for intelligent agricultural machinery path planning, injecting new impetus into the development of smart agriculture. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1It is a flow chart of the method for planning the operation path of a corn harvester in the field for reducing mechanical soil compaction according to the present invention;

[0032] Figure 2 It is the flowchart of the Gray Wolf optimization algorithm;

[0033] Figure 3 is the flow chart of the cuckoo search algorithm;

[0034] Figure 4 is the flow chart of the immune algorithm;

[0035] Figure 5 is the schematic diagram of an I-turn, where Figure 5 (a) and Figure 5 (b) is a schematic diagram of the vehicle traveling to different locations;

[0036] Figure 6 is the schematic diagram of Type II turning, where Figure 6 (a) and Figure 6 (b) is a schematic diagram of the vehicle traveling to different locations. DETAILED DESCRIPTION

[0037] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0038] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0039] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments, but they are not intended to limit the present invention.

[0040] Specific implementation method 1: The following is combined Figures 1 to 6 The present embodiment describes a method for planning a corn harvester's field operation path for reducing soil mechanical compaction. Figure 1 , the method comprises the following steps:

[0041] Step 1: Obtain GNSS-based in-field ridge and ditch information and boundary point information of the farmland to be planned;

[0042] Step 2: Arrange sensors;

[0043] Step 3: Determine whether the farmland to be planned is a regular field. If it is a regular field, execute step 4; if it is not a regular field, execute step 6;

[0044] Step 4: Construct a path planning model for regular fields;

[0045] Step 5: Use the Gray Wolf Optimization Algorithm to solve the path planning model in the regular field and output the optimal path for the regular field operation;

[0046] Step 6: Determine whether the turning radius is less than the maximum turning radius allowed by the plot. If so, execute the Type I turning method, and then execute Step 7; if not, execute the Type II turning method, and then execute Step 8;

[0047] Step 7: Use the cuckoo search algorithm to obtain the optimal turning path parameters under the I-type turning method, and then execute step 9;

[0048] Step 8: Using the cuckoo search algorithm to obtain the optimal turning path parameters under the type II turning method;

[0049] Step 9: Use the immune algorithm to plan the path for irregular fields and output the optimal path for irregular field operations.

[0050] In step 1, GNSS (Global Navigation Satellite System) is used to determine the boundary and furrow information. The total number of furrows in the field to be operated is N.

[0051] In step 2, sensors are arranged in the ridges, S is the sensor arrangement scheme, (j, a) means that a sensor is arranged at the jth ridge a, and the sensor is used to detect whether it is compressed. The purpose of detecting this information is to design an optimized path to compact the same position as little as possible and reduce the degree of soil compaction.

[0052] In step three, the paths for regular plots and irregular plots are planned in different ways, and the present invention plans them separately.

[0053] In step 4, a path planning model for regular field plots is constructed:

[0054] Objective Function In the formula, α, β are the weight coefficients of the objective function;

[0055] y i Indicates whether the variable is harvested.

[0056] z i,j Indicates whether the ridge pair (i, j) is compacted variable,

[0057] Decision variables: represents the path of the kth corn harvester, where p k,trepresents the ridge pair of the k-th corn harvester's t-th corn harvester path, t∈{1,2,...,T k}, T k represents the maximum number of operation steps of each corn harvester; each p k,t ={i,j} represents the two ridges compacted in the current operation;

[0058] Constraints:

[0059] Condition 1: Harvest coverage constraints: x k,t,i Indicates whether the k-th corn harvester harvests the i-th ridge at the t-th harvest. Each ridge must be harvested once, and N is the total number of ridges in the field;

[0060] Condition 2: Compaction minimization constraint: x k,t,j Indicates whether the k-th corn harvester compacts the sensor of the j-th ridge during the t-th harvest, limiting the number of compactions of the sensor of each ridge to at most once, and S is the sensor layout plan;

[0061] Fixed compaction ridge constraint: Among them, K is the harvester set, and the corn harvester only compacts two ridges each time it travels;

[0062] A ridge constraint that spans but does not compact: Corn harvester can cross R on the path k Make ridges, but compact only 2 of them.

[0063] In step 5, the gray wolf optimization algorithm is used to solve the path planning model of the regular field, and the optimal path of the regular field operation is output; the specific model solving process is:

[0064] (1) Initialization: population size P, each individual P k Represents a path plan and calculates the fitness value f of each individual path path (P k ); (2) Hierarchical division: c Gray wolf: current optimal path P c . g Gray Wolf: Suboptimal Path P g . δGray Wolf: The third best path P δ ; (3) Path update: Calculate the distance D between other path individuals in the population and the c, g and δ paths c =|C1·P c -P k |,D d =|C2·P g -P k |,D δ =|C3·P δ -P k|, adjust the path according to the distance Among them A m ′,C m is a dynamic adjustment coefficient, m = 1, 2, 3, used to balance global search and local development; (4) Fitness calculation and update: for each updated path P k , recalculate the fitness value f path (P k ). Update c, g, δ and the gray wolf path; (5) Iterate and output the optimal path: Repeat the path update and fitness calculation until the maximum number of iterations T is reached or the path converges, and output the gray wolf path P of c c as the optimal path.

[0065] In step 6, the turning method in the field under irregular fields is based on the Ackerman steering principle. The general formula of Ackerman steering is: Where R is the turning radius; L is the distance between the front and rear wheels of the corn harvester; θ is the steering angle of the front wheel. max When θ max Indicates the maximum steering angle of the harvester's front wheels; when the turning radius R is greater than the maximum turning radius R allowed by the plot max When there is enough space on the plot. Applicable to regular paths under ample conditions, and Type II turns θ min Indicates the minimum steering angle of the harvester's front wheels (or the minimum steering limit in some special circumstances).

[0066] See also Figure 5 As shown, the turning radius R is less than the maximum turning radius R allowed by the plot. max When the I-type turning method is used, the turning path parameters h′, r′, d′, and the path length Wherein, h' is the projection of the distance between the starting and ending points of the planned path in the I-type turning method in the planting direction of the field ridge, r' is the radius of the virtual turning path in the I-type turning method, d' is the distance between the field ridges inside the field in the I-type turning method, and e is the center distance between the front and rear tire spindles of the corn harvester;

[0067] See also Figure 6 As shown, the turning radius R in step 6 is greater than the maximum turning radius R allowed by the plot. max When the turning method is used, the type II turning method is used. The type II turning method requires the engine to be reversed twice in one turning operation, that is, it is necessary to complete one "moving (turning)-reversing (straight line)-moving (turning)" operation. The turning path parameters are h", r, d, and the path length Occupied land area Wherein, h″ is the projection of the distance between the starting and ending points of the planned path in the type II turning method in the planting direction of the field ridge, r″ is the radius of the virtual turning path in the type II turning method, d″ is the distance between the field ridges inside the field in the type II turning method, W is the vehicle's outer width, and W′ is the vehicle's tire width.

[0068] See also Figure 3 In step 7, the cuckoo search algorithm is used to obtain the optimal turning path parameters under the I-type turning method. The specific process is as follows: parameter setting: N′ population size (number of bird nests); α′ step size factor, set to 0.1; P a The probability of being replaced is 0.25, Τ max Maximum number of iterations; r m ' in ,r m ' ax ,h′ min ,h′ max ,d′ min ,d′ max Variable range; e is the center distance of the main axles of the front and rear tires of the corn harvester. Population initialization: Randomly generate N' solutions, each of which is the turning path parameter x i =[h i ′,r i ′,d i ′],h i ′∈[h′ min ,h′ max ],r i ′∈[r m ' in ,r m ' ax ],d i ′∈[d′ min ,d′ max ], for each solution x i , calculate the objective function value And record the optimal solution and its target value x best , Whether the termination condition is met: If 1. The maximum number of iterations t = T is reached max ; 2. If there is no significant change in the optimal solution for multiple generations (i.e., the algorithm converges), the algorithm is terminated and the optimal solution is output. Generate candidate population operator: For each solution x in the population i , using Levy flight to generate new candidate solutions x i ′:x i ′=x i +α′·Le′vy(λ), where Le′vy(λ) represents the random step size of the Levy distribution, σ 2 is the scale parameter of the distribution: λ = 1.5. Select the best operator: Calculate each solution x i The objective function value F(x i ′), compared with the previous solution: Random migration operator: with probability P a Randomly select a portion of the solutions in the population and replace them with new solutions (generated by random reinitialization): j =[h j ′,r j ′,d j ′],h j ′∈[h′ min ,h′ max ],r j ′∈[r m ' in ,r m ' ax ],d j ′∈[d′ min ,d′ max ], after replacement, recalculate the objective function value F(x i ). Select the best operator again: Update the optimal value x again best , F best The algorithm ends and outputs the optimal solution: optimal path parameter x best =[h * ,r * ,d * ], the optimal path length l * =F(x best ).

[0069] See also Figure 3 In step eight, the cuckoo search algorithm is used to obtain the optimal turning path parameters under the type II turning method; the specific process is: construct a multi-objective function: minJ = ω1l′ + ω2S′, where ω1 and ω2 are weight coefficients used to adjust the trade-off between path length and occupied area. Parameter settings: N″ population size (number of cuckoo nests); α″ step size factor, controlling the jump range of the new solution (α″ = 0.01); P a ′The probability of being abandoned (recommended value P a ′=0.25); T m ' ax Maximum number of iterations; the range of h″, r″, d″ is h″∈[h′ m ' in ,h′ m ' ax ],r″∈[r m ″ in ,r m ″ ax ],d″∈[d′ m 'in ,d′ m ' ax ]. Population initialization: randomly generate N″ solutions, defined as x i ″=[h i ″,r i ″,d i ″] and randomly distributed within the parameter range: h i ″~U(h″ min ,h″ max ),r i ″~U(r″ min ,r″ max ),d″ i ~U(max(d″ min ,2r i ″),d″ max ). Target value calculation: J(x i ″)=ω1l′(x i ″)+ω2S′(x i ″) Check whether the termination condition is met. If T m ' ax , the algorithm is terminated and the current optimal solution is output. Generate candidate population operator: Generate a new candidate solution x for each cuckoo nest i ″′:x i ″′=x i ″+α″·L′(s,λ′), α″ is the step size factor, L(s,λ′), the random step size that follows the Le′vy flight distribution λ′=1.5, for the generated new solution x i ″′, calculate the target value J(x i ″′), and keep the better solution using the following rules Random migration operator: with probability P a ′ replaces part of the solution with a random new solution, x i ″=x new , if U(0,1)<P a , x new is the newly generated solution. Select the operator again: for the current population, calculate the target value J(x i ″), retain the optimal solution: Check the termination condition and repeat the above steps until the maximum number of iterations Τ′ is reached max Or the change of the solution is less than the preset threshold. Algorithm output: including the optimal path parameters [h *′ ,r *′ ,d *′ ]; optimal path length l *′ and the minimum headland area S *′ ; Optimize function value J *′ .

[0070] See also Figure 4 In step nine, the immune algorithm is used to plan the path for the irregular field and output the optimal path for the irregular field operation. The specific process of the path rule in the irregular farmland is as follows:

[0071] (1) Initialize the field operation sequence: Generate a random operation sequence `T = {`T(1), `T(2), ..., `T(N)}, where `T(i) represents the operation sequence of the i-th ridge, N is the total number of ridges, and for each corn harvester of different sizes, specify the number of ridges it crosses `ω. Each trip will cross `ω ridges but only compact two ridges; (2) Calculate affinity: Define the affinity function `F(A i ″) for each antibody A i The fitness value of ″ is: Where `L(`T(j),`T(j+1)) represents the length of the working path; `C(`T(j),`T(j+1)) is the penalty term for ridge compaction, which is determined by the size of the corn harvester. If the corn harvester crosses `ω ridges but only compacts two, the number of ridges crossed is calculated and the remaining ridges are avoided from compaction; P(`T) is the total compaction of the sensors, which represents the number of sensors passed; `α and γ are corresponding weight parameters used to control the impact of compaction penalty and sensor compaction on fitness; (3) Check termination conditions: If the maximum number of iterations or fitness convergence conditions are met, the current optimal path is output; otherwise, the iteration continues; (4) Immune selection: The affinity `F(A i ″) for roulette wheel selection, the antibody with higher fitness has a greater selection probability. M is the total number of antibodies in the population; (5) Cloning: According to the antibody affinity, perform appropriate cloning, assuming the number of clones is n i , n i =round(clone_rate×F(A i )), clone_rate is the clone ratio parameter; (6) Mutation: In each cloned antibody, the operation order of the two ridges is randomly exchanged. i and `T j, adjust the path and adapt to the size of the corn harvester, update the transfer paths of adjacent ridges `L(`T(i-1),`T(i)) and `L(`T(j),`T(j+1)), ensure that the number of ridges compacted by the corn harvester each time meets its width limit, and only compact two ridges. T″′(i)=`T(j),T″′(j)=`T(i); (7) Compaction judgment: add the compaction constraint function g(`T) to ensure that the number of compaction sensors P(`T) is minimized: minP(`T)subject to g(`T)≤compactment threshold; (8) Record the total transfer path and perform clone suppression: add the solutions that meet the constraints to the solution set, apply clone suppression, reduce the duplicates in the solution set to maintain diversity; (9) Update the population: refresh the population, retain the high fitness solution, and generate a new initial solution to form the population for the next iteration; (10) Repeat iteration: iteratively update the population and gradually optimize the solution until the termination condition is met.

[0072] Although the present invention is described herein with reference to specific embodiments, it should be understood that these embodiments are merely examples of the principles and applications of the present invention. It should therefore be understood that many modifications may be made to the exemplary embodiments and that other arrangements may be devised without departing from the spirit and scope of the present invention as defined by the appended claims. It should be understood that the various dependent claims and features described herein may be combined in a manner different from that described in the original claims. It should also be understood that the features described in conjunction with a single embodiment may be used in other described embodiments.

Claims

1. A method for planning a corn harvester's field operation path for reducing soil mechanical compaction, characterized in that: The method comprises the following steps: Step 1: Obtain GNSS-based in-field ridge and ditch information and boundary point information of the farmland to be planned; Step 2: Arrange sensors; Step 3: Determine whether the farmland to be planned is a regular field. If it is a regular field, execute step 4; if it is not a regular field, execute step 6; Step 4: Construct a path planning model for regular field plots; Step 5: Use the Gray Wolf Optimization Algorithm to solve the path planning model in the regular field and output the optimal path for the regular field operation; Step 6: Determine whether the turning radius is less than the maximum turning radius allowed by the plot. If so, execute the Type I turning method, and then execute Step 7; if not, execute the Type II turning method, and then execute Step 8; Step 7: Use the cuckoo search algorithm to obtain the optimal turning path parameters under the I-type turning method, and then execute step 9; Step 8: Using the cuckoo search algorithm to obtain the optimal turning path parameters under the type II turning method; Step 9: Use the immune algorithm to plan the path for irregular fields and output the optimal path for irregular field operations.

2. The method for planning a corn harvester field operation path for reducing soil mechanical compaction according to claim 1, characterized in that: The construction of the path planning model for regular field plots in step 4 is as follows: Objective Function In the formula, α, β are the weight coefficients of the objective function; y i Indicates whether the variable is harvested. z i,j Indicates whether the ridge pair (i, j) is compacted variable, Decision variable: P k ={p k,1 ,p k,2 ,...,p k,Tk } represents the path of the kth corn harvester, where p k,t represents the ridge pair of the k-th corn harvester's t-th corn harvester path, t∈{1,2,...,T k }, T k represents the maximum number of operation steps of each corn harvester; each p k,t ={i,j} represents the two ridges compacted in the current operation; Constraints: Condition 1: Harvest coverage constraints: x k,t,i Indicates whether the k-th corn harvester harvests the i-th ridge at the t-th harvest. Each ridge must be harvested once, and N is the total number of ridges in the field; Condition 2: Compaction minimization constraint: x k,t,j Indicates whether the k-th corn harvester compacts the sensor of the j-th ridge during the t-th harvest, limiting the number of compactions of the sensor of each ridge to at most once, and S is the sensor layout plan; Fixed compaction ridge constraint: Among them, K is the harvester set, and the corn harvester only compacts two ridges each time it travels; A ridge constraint that spans but does not compact: Corn harvester can cross R on the path k Make ridges, but compact only 2 of them.

3. The method for planning a corn harvester field operation path for reducing soil mechanical compaction according to claim 2, characterized in that: Step 5: Take the path P of the kth corn harvester k As the initial population individual of the gray wolf optimization algorithm, it iteratively outputs the optimal path for regular field operations.

4. The method for planning a corn harvester field operation path for reducing soil mechanical compaction according to claim 1, characterized in that: The turning radius R in step 6 is less than the maximum turning radius R allowed by the plot. max When the I-type turning method is used, the turning path parameters h′, r′, d′, and the path length Wherein, h' is the projection of the distance between the starting and ending points of the planned path in the I-type turning method in the planting direction of the field ridge, r' is the radius of the virtual turning path in the I-type turning method, d' is the distance between the field ridges inside the field in the I-type turning method, and e is the center distance between the front and rear tire spindles of the corn harvester; The turning radius R in step 6 is greater than the maximum turning radius R allowed by the plot. max When the type II turning method is used, the turning path parameters are h″, r″, d″, and the path length Occupied land area Where h″ is the projection of the distance between the starting and ending points of the planned path in the type II turning method in the planting direction of the ridges, r″ is the radius of the virtual turning path in the type II turning method, and d″ is the distance between the ridges inside the field in the type II turning method.

5. The method for planning a corn harvester field operation path for reducing soil mechanical compaction according to claim 1, characterized in that: In step nine, the immune algorithm is used to plan the path for the irregular field, and the operation sequence is used as the population individual. The operation sequence is the path of the corn harvester, and the optimal path for the irregular field operation is output through iteration.

Images