Automatic control system of electric agricultural machine

Abstract

The invention discloses an automatic control system of an electric agricultural machine, and relates to the technical field of control systems. The system comprises an atomic task decomposition and energy consumption modeling module, which is used for decomposing a macroscopic job into three types of atomic tasks of movement, job and action and establishing a refined energy consumption model; the global sequence optimization module is used for optimizing an atomic task sequence and distributing an energy budget by taking total energy as a constraint and taking comprehensive benefit maximization as a target; the real-time energy consumption monitoring and budget cancel-after-verification module is used for tracking execution and accounting energy consumption deviation; and the dynamic re-planning decision module triggers the re-optimization of the residual atomic task parameters or sequences when the deviation threatens the total atomic task. Through atomic task level energy budget management, global energy optimal planning before operation and dynamic adaptive adjustment in operation are realized, the problems of low operation reliability and low energy utilization efficiency caused by limited electric quantity of the electric agricultural machinery are solved, and the effective operation quantity of single charging is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of control systems, in particular to an automatic control system for electric agricultural machinery. BACKGROUND

[0002] With the deepening of electrification and intelligentization in the field of agriculture, electric tractors, electric harvesters and other electric agricultural machinery have become a development trend. However, electric agricultural machinery faces a core bottleneck: the sharp contradiction between limited on-board battery energy and complex and dynamic field operation atomic tasks. Specifically, 1) it is impossible to generate an energy-optimal global operation sequence before operation according to specific field maps and operation atomic tasks, resulting in "calculate as you go", which often causes improper energy allocation in the early stage, and the key operation cannot be completed or is forced to return to charge; 2) complex operations such as "plowing" are regarded as a homogeneous energy consumption process, ignoring the huge energy consumption difference between sub-atomic tasks such as "lifting plow-falling plow-straight plowing-turning at the end of the field", resulting in serious distortion of real-time energy estimation; 3) when encountering unexpected local high ground resistance and soft mud area during operation, the existing system can only passively increase the motor torque, and cannot dynamically adjust the operation sequence or parameters of subsequent unexecuted atomic tasks to compensate for the additional energy consumption and ensure the completion of the total atomic task. SUMMARY

[0003] The present application aims to solve the above-mentioned problems of the prior art and provide an automatic control system for electric agricultural machinery which can perform global energy optimization planning before operation, finely perceive the energy consumption of heterogeneous atomic tasks during operation and dynamically re-plan, so as to maximize the effective operation amount per charge and improve operation reliability and user predictability.

[0004] The specific scheme is as follows: An automatic control system for electric agricultural machinery, comprising an atomic task decomposition and energy consumption modeling module, a global sequence optimization module, a real-time energy consumption monitoring and budget verification module, a dynamic re-planning decision module and a human-computer interaction and execution interface module, which are connected in sequence; The atomic task decomposition and energy consumption modeling module is used to decompose the macro field operation atomic task input by the user into a plurality of atomic tasks with different energy consumption characteristics, and establish a basic energy consumption model for each type of atomic task; the global sequence optimization module is used to optimize the execution sequence of the atomic tasks under the total energy constraint, and allocate energy budget for each atomic task in the optimized operation sequence, with the goal of maximizing the overall benefit of operation; The real-time energy consumption monitoring and budget verification module is used to track the currently executed atomic task and accumulate its actual energy consumption during the operation of the agricultural machinery, and calculate the deviation between the actual energy consumption and the allocated energy budget; The dynamic re-planning decision module is configured to trigger re-optimization of the execution sequence or parameters of the remaining unexecuted atomic tasks when the deviation exceeds a threshold and the evaluation will affect the overall atomic task completion, and generate an updated energy budget and control instruction.

[0005] Further, the atomic task decomposition and energy consumption modeling module comprises a map and atomic task parsing submodule configured to parse a digital map containing fields, obstacles and charging station locations and a user-specified work type and range; An atomic task generation submodule is configured to decompose a macro work atomic task into a standardized atomic task sequence based on the parsing results, the atomic task comprising at least three types of atomic tasks, i.e., a movement atomic task, a work atomic task and a mechanism action atomic task; A heterogeneous energy consumption model library submodule is configured to store energy consumption calculation formulas and basic parameters corresponding to the three types of atomic tasks.

[0006] Further, for a straight line movement type atomic task, the heterogeneous energy consumption model library submodule estimates the movement atomic task energy consumption E move as follows: E move =(f roll ·m·g·cosθ+m·g·sinθ+1 / 2·ρ·C d ·A·v 2 )·L / η; wherein f roll is a rolling resistance coefficient, m is the mass of the agricultural machine, g is the acceleration of gravity, θ is the slope angle, ρ is the air density, C d is the wind resistance coefficient, A is the windward area, v is the set speed, L is the movement distance, and η is the transmission efficiency.

[0007] Further, for a field work type atomic task, the heterogeneous energy consumption model library submodule estimates the work atomic task energy consumption E work as follows: ; wherein is the basic working resistance of the agricultural implement, H soil is the soil hardness coefficient, k is the proportional coefficient, L work is the work travel distance, and η is the transmission efficiency. Further, for a mechanism action type atomic task, the heterogeneous energy consumption model library submodule estimates the mechanism action atomic task energy consumption E action as follows: E action =P hydraulic τ action ; wherein P hydraulicThe hydraulic system rated power, τ action is the action duration.

[0008] Further, the global sequence optimization module comprises a constraint management submodule, an optimization algorithm engine submodule and a sequence and budget output submodule. The constraint management submodule is configured to define and manage total energy constraints, inter-atomic task sequence constraints and agricultural machinery kinematics constraints. The optimization algorithm engine submodule is configured to take the maximum comprehensive benefit of the operation as an objective function, take the total energy constraints as core constraints, perform search calculation on the sequence of atomic tasks, and output at least one candidate optimized operation sequence. The sequence and budget output submodule is configured to allocate an energy budget based on the estimated energy consumption of each atomic task in the selected optimized operation sequence.

[0009] Further, the optimization algorithm engine submodule adopts a genetic algorithm for heuristic search, and the comprehensive benefit of the operation U(S) is represented as: U(S) = w area ·A(S) / A total +w time ·T ideal / T(S). Wherein, A(S) is the effective operation area completed by the sequence (considering the geometric calculation of path coverage); T(S) is the estimated total time consumption of the sequence; w area and w time are user-set weights, and w area +w time =1; A total and T ideal are used for normalization; A total refers to the total target atomic task area set by the user, which is the benchmark value of operation benefit; T ideal refers to the shortest theoretical time required to complete all atomic tasks under ideal working conditions without considering energy constraints and at the highest efficiency, which is the normalized benchmark of time benefit; the operation area benefit A(S) / A total represents the proportion of the area completed by the sequence to the total area; the operation time efficiency benefit T ideal / T(S) represents the ratio of the ideal time consumption T ideal to the actual estimated time consumption T(S) (the larger the value, the faster the operation speed); Wherein, L i is the length (distance) of the atomic task i, v i is the speed of performing the atomic task, L i / v i is the theoretical time consumption of the atomic task, and T turnThe total accumulated time consumption of all non-traveling fixed actions such as turning at the end of a sequence in the sequence.

[0010] Further, the real-time energy consumption monitoring and budget cancellation module comprises a multi-source sensor fusion submodule, an atomic task execution tracking submodule, and an energy consumption cancellation and deviation calculation submodule. The multi-source sensor fusion submodule is configured to collect battery electrical parameters, motor parameters, vehicle position, and working mechanism state data in real time. The atomic task execution tracking submodule is configured to determine the atomic task currently being executed according to the vehicle position and working mechanism state data. The energy consumption cancellation and deviation calculation submodule is configured to accumulate the actual energy consumption of the current atomic task and calculate the deviation value ΔE of the actual energy consumption from the corresponding energy budget.

[0011] Further, the dynamic re-planning decision module comprises an influence evaluation submodule, a re-planning trigger and strategy selection submodule, and a local sequence regeneration submodule. The influence evaluation submodule is configured to evaluate the influence on the total energy budget of the remaining atomic tasks based on the energy consumption deviation of the current atomic task. The re-planning trigger and strategy selection submodule is configured to trigger re-planning when the ratio of the remaining total budget to the total estimated energy consumption of the remaining atomic tasks is lower than a preset threshold, and to select one of a parameter adjustment strategy, a sequence adjustment strategy, or an inserted charging strategy. The local sequence regeneration submodule is configured to call the global sequence optimization module to re-optimize the remaining unexecuted atomic tasks according to the selected strategy.

[0012] Further, the parameter adjustment strategy refers to reducing the working speed or working depth of subsequent non-critical atomic tasks; the sequence adjustment strategy refers to skipping lower-priority working atomic tasks; and the inserted charging strategy refers to planning a path to the nearest charging station and inserting it into the current working sequence.

[0013] Compared with the prior art, the present application has at least one of the following technical effects: 1. The present application solves the problem of the lack of global energy planning in the prior art, and realizes optimal energy allocation before operation. Through the collaborative work of the "atomic task decomposition and energy consumption modeling module" and the "global sequence optimization module", the present application solves the problem of blindness of "calculating where you go". Specifically, the system first decomposes macro atomic tasks such as "plowing field A" into a series of atomic tasks such as "empty movement", "lowering plow", "straight plowing", "raising plow", and "turning at the end of the field" based on map analysis and coverage path algorithm. Then, the built-in heterogeneous energy consumption model is used to calculate the estimated energy consumption of each type of atomic task based on physical principles. On this basis, the "global sequence optimization module" takes the total battery energy as the core constraint, and maximizes the comprehensive benefits of operation area and timeliness as the goal, and uses genetic algorithm and other methods to globally search and optimize the atomic task sequence. This makes the system be able to plan an optimal energy utilization and agronomic logic operation path (such as using spiral plowing from the center to reduce empty running) before the operation starts, and allocate accurate "energy budget" for each atomic task. Thus, it ensures that high-priority atomic tasks can be completed, and avoids energy depletion or forced interruption of operation due to improper early planning.

[0014] 2. The present application solves the problem of coarse granularity of energy consumption perception in the prior art, and realizes fine and differentiated energy consumption estimation and management. The present application abandons the rough method of estimating energy consumption as a whole, and realizes fine perception of energy consumption heterogeneity within the operation through "atomic task decomposition" and "heterogeneous energy consumption model library". The key technology is that the system standardizes complex operation processes into atomic tasks with three different energy consumption characteristics: movement, operation, and mechanism action, and establishes independent physical models for each type of atomic task. For example, for "straight plowing", its energy consumption model not only considers the distance and speed, but also introduces the soil hardness variable, so that the energy consumption estimation can reflect the soil quality difference within the field; for "turning at the end of the field", its energy consumption may consist of fixed hydraulic action power consumption and low-speed large-torque driving power consumption. This differentiated modeling based on atomic task type changes the total energy consumption estimation from "one fuzzy total value" to "the sum of a series of accurate itemized budgets", improving the accuracy and reliability of budget preparation, and laying a solid scientific foundation for subsequent optimization and monitoring.

[0015] 3、Solve the problem of insufficient dynamic response capability in the prior art, and realize self-adaptive energy re-planning in the operation process. The present application constructs a complete "monitoring-decision-adjustment" closed loop through the "real-time energy consumption monitoring and budget verification module" and the "dynamic re-planning decision module", and gives the system strong robustness to respond to field uncertainties. In specific implementation, the system fuses battery power, GNSS position and agricultural machine state data in real time during operation, accurately tracks the currently executed atomic task, and compares the energy consumption calculated by integration with the budget. Once it is found that a certain atomic task (such as a section of tillage) is over-standard in energy consumption due to abnormally hard soil and threatens the completion of the overall atomic task, the "dynamic re-planning decision module" will start immediately. It will evaluate the inequality between the remaining energy and the energy required by the atomic task, and intelligently select parameter adjustment (such as automatically reducing the speed of subsequent tillage), sequence adjustment (such as skipping the low-priority area at the edge of the field) or insertion of charging, and call the optimization engine to quickly recalculate the subsequent optimal scheme. This makes the system change from passively bearing energy consumption impact to actively managing energy crisis, ensuring that the core operation atomic task can still be completed with the maximum remaining power in the event of an emergency, and improving the reliability and success rate of the operation. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description only constitute some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without any creative effort.

[0017] Figure 1 is a module structure diagram of an automatic control system of an electric agricultural machine. DETAILED DESCRIPTION

[0018] In the following description, specific details such as specific system structures, techniques, etc. are presented in order to thoroughly understand the embodiments of the present application. However, it should be clear to those skilled in the art that the present application can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits and methods are omitted to avoid unnecessary details that hinder the description of the present application.

[0019] It should be understood that when used in the specification and the appended claims of the present application, the term "comprising" indicates the presence of the described features, whole, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, whole, steps, operations, elements, components and / or sets thereof.

[0020] It should also be understood that the term "and / or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items, and that the term "at least one of' as used herein means "one or more" of the listed items, that the term "permutation" as used herein means "all possible combinations", and that the term "coupled" as used herein means "directly or indirectly coupled".

[0021] As used in the description of the application and the appended claims, the term "if' can be interpreted to mean "when" or "upon" or "in response to determining" or "in response to detecting" depending on the context. Similarly, the phrase "if it is determined" or "if [a described condition or event] is detected" can be interpreted to mean "upon determining" or "in response to determining" or "upon [the described condition or event] being detected" or "in response to [the described condition or event] being detected", depending on the context.

[0022] In addition, in the description and drawings of the present application, the terms "first", "second", "third", etc. are used only for distinguishing between similar objects, and cannot be interpreted as implying or suggesting relative importance.

[0023] Reference throughout this specification to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the application. Thus, the appearances of the phrases "in one embodiment", "in some embodiments", "in other embodiments", "in additional embodiments", and the like, in various places throughout this specification are not necessarily all referring to the same embodiment, unless otherwise specifically indicated. The terms "comprising", "comprises", "including", "includes" and "having" and their variants are meant to encompass the terms "including" and "including but not limited to", unless otherwise specifically indicated. The present application proposes a new mode of "atomic task level energy budget management", aiming to solve the core problem of low reliability of atomic task completion and low efficiency of energy utilization of electric agricultural machines due to limited battery energy.

[0024] Specifically, referring to the accompanying drawings Figure 1 , an automatic control system for electric agricultural machines is disclosed, which comprises an atomic task decomposition and energy consumption modeling module, a global sequence optimization module, a real-time energy consumption monitoring and budget verification module, a dynamic re-planning decision module, and a human-computer interaction and execution interface module, and the modules are connected to each other; The atomic task decomposition and energy consumption modeling module is used to decompose the macroscopic field operation atomic task input by the user into a plurality of atomic tasks with different energy consumption characteristics, and to establish a basic energy consumption model for each type of atomic task; the global sequence optimization module is used to optimize the execution sequence of the atomic tasks under the total energy constraint, with the goal of maximizing the overall efficiency of the operation, and to allocate energy budget for each atomic task in the optimized operation sequence; The real-time energy consumption monitoring and budget verification module is configured to track the current executed atomic task and accumulate its actual energy consumption, and calculate the deviation of the actual energy consumption from the allocated energy budget during the operation of the agricultural machine. The dynamic re-planning decision module is configured to trigger the re-optimization of the execution sequence or parameters of the remaining unexecuted atomic tasks, and generate an updated energy budget and control instruction, when the deviation exceeds a threshold and the evaluation will affect the completion of the overall atomic task. The human-computer interaction and execution interface module is configured to show the user the operation sequence and energy budget, and convert the atomic task instruction into a bottom-layer control signal. The automatic control system of the electric agricultural machine runs on a vehicle-mounted computing platform, and the working flow is as follows: the atomic task decomposition and energy consumption modeling module receives the operation instruction, and outputs an atomic task list with an energy consumption model; the global sequence optimization module calculates the energy-optimal operation sequence and budget according to the atomic task list; the human-computer interaction and execution interface module guides the start of the operation, and sends the atomic task instruction to the vehicle controller; the real-time energy consumption monitoring and budget verification module monitors the deviation of the actual energy consumption from the budget during the operation; when the deviation threatens the overall atomic task, the dynamic re-planning decision module intervenes, adjusts the subsequent plan, and updates the instruction. The modules are loosely coupled through well-defined data interfaces (such as structured messages such as atomic task list, optimized operation sequence, and energy consumption alert).

[0025] In specific implementation, the atomic task decomposition and energy consumption modeling module includes: a map and atomic task parsing sub-module configured to parse a digital map containing a field block, an obstacle, and a charging station location, and a user-specified operation type and range; by inputting standard geographic information data (such as a KML / Shapefile file of a farmland boundary) and a user-selected operation type (an enumerated value such as “PLOUGHING” and “SEEDING”), etc., the geographic information system library is used to parse the boundary coordinates, convert them into internal unified plane coordinates (such as UTM), and construct a geometric polygon describing the field block area, and at the same time, load the preset charging station and obstacle point information.

[0026] An atomic task generation sub-module is configured to decompose the macroscopic operation atomic task into a standardized atomic task sequence based on the parsing result, and the atomic task includes at least three types of moving atomic task, operation atomic task, and mechanism action atomic task. Specifically, it includes: Path planning: based on the agricultural machine width W, a “bow-shaped” or “spiral-shaped” coverage path algorithm is used to generate a continuous path point sequence P = [P0, P1, P2...P n ], wherein P i is a two-dimensional coordinate.

[0027] Atomic task semantic segmentation: traverse the path point sequence, and cut the continuous path into atomic tasks according to rules. For example, the path segment connecting the start and end points of two work rows is identified as a “turning at the end of the work row atomic task”; the path segment on the same work row is identified as a “straight work atomic task”, and the length L = ‖P i+1 -P i ‖; at the beginning and end of the work atomic task, automatically insert “farm implement lowering” and “farm implement lifting” mechanism action atomic tasks; the movement from the initial position to the first work point is identified as an “empty load movement atomic task”. The heterogeneous energy consumption model library submodule is used to store the energy consumption calculation formulas and basic parameters corresponding to the above three types of atomic tasks. The structure of the heterogeneous energy consumption model library submodule is a queryable database or configuration file, which stores three types of models: Model index: atomic task type → calculation formula identifier; Formula parameters: basic parameters required in the calculation formula (such as farm machine mass m, rolling resistance coefficient f roll , farm implement basic resistance F implement , etc.).

[0028] Environment parameter interface: pointing to sensors or data sources for obtaining dynamic parameters (such as real-time slope θ, soil hardness H soil ).

[0029] In specific implementation, for the straight movement type atomic task, the calculation formula of the heterogeneous energy consumption model library submodule for estimating the energy consumption E move of the movement atomic task is: E move =(f roll ·m·g·cosθ+m·g·sinθ+1 / 2·ρ·C d ·A·v 2 )·L / η; where f roll is the rolling resistance coefficient, m is the mass of the farm machine, g is the acceleration of gravity, θ is the slope angle, ρ is the air density, C d is the wind resistance coefficient, A is the windward area, v is the set speed, L is the movement distance, and η is the transmission efficiency. This formula is a direct application of the theory of vehicle driving resistance. The total resistance F total is composed of three parts: F total =F roll +F grade +F aero ; where F roll =f roll ·m·g·cosθ is the rolling resistance; F grade =m·g·sinθ is the slope resistance; F aero =1 / 2·ρ·C d ·A·v2 The work done to overcome the total air resistance and travel a distance L is F. total ·L, considering the transmission efficiency η, the battery output energy is E move =F total ·L / η. Where θ is calculated based on the digital elevation model data of the path points. The digital elevation model (DEM) is a standard geographic data product that provides terrain elevation information and is common knowledge in precision agriculture and autonomous driving path planning of agricultural machinery in this field; v is the cruising speed set for this atomic task; the others are inherent vehicle parameters.

[0030] In practical implementation, for field operation-type atomic tasks, the heterogeneous energy consumption model library submodule estimates the energy consumption E of the operation atomic task. work The calculation formula is: ; in, For the basic working resistance of agricultural machinery, H soil L is the soil hardness coefficient, k is the proportionality coefficient, and L is the soil hardness coefficient. work Let η be the distance traveled during operations, and η be the transmission efficiency. The energy consumption model for field operations simplifies the operating resistance into a force linearly related to soil hardness. "Basic working resistance of agricultural machinery" is a fixed component of the total "operating resistance" that does not change with soil hardness and is determined by the structure of the agricultural machinery itself and the set operating depth. The relationship between the two is: operating resistance = basic working resistance + k × soil hardness, where k was obtained through traction tests of agricultural machinery in standard soil. The energy consumption calculation for atomic tasks in field operations is similar to that of the movement model; the operating energy consumption is the work done to overcome operating resistance divided by the transmission efficiency E of the operating mechanism. work =(F work ·L work ) / η。 H soil It can be obtained through one of the following methods: 1) a pre-drawn soil hardness map; 2) real-time estimation and location binding by an airborne near-infrared sensor; 3) online calibration based on motor torque feedback from the first few rows of operation.

[0031] In specific implementation, for atomic tasks involving mechanism actions, the heterogeneous energy consumption model library submodule estimates the energy consumption E of the atomic task involving mechanism actions. action The calculation formula is: E action =P hydraulic τ action ; Where P hydraulic τ is the rated power of the hydraulic system. actionFor the duration of the action, the energy consumption of an atomic task of mechanism action type is the product of the rated power of the hydraulic system and the duration of the action.

[0032] In specific implementation, the global sequence optimization module includes a constraint management submodule, an optimization algorithm engine submodule, and a sequence and budget output submodule. The global sequence optimization module formalizes the agricultural operation planning problem into a constrained combinatorial optimization problem and provides specific solution ideas and objective functions, which is the core of the "global energy optimal planning" algorithm.

[0033] The constraint management submodule is used to define and manage total energy constraints, atomic task sequence constraints, and agricultural machinery kinematic constraints; among which, energy constraints include: Where r is the safety factor (e.g., 0.9), E battery This represents the current available power. This represents the estimated energy consumption of all atomic tasks in the optimized job sequence S. The sum; the overall meaning of energy constraint is: the estimated energy consumption of all atomic tasks in the optimization job sequence S. The sum of these must not exceed the current available battery capacity E. battery The available budget, multiplied by a safety factor γ, is the core mathematical constraint for achieving global energy management. Sequence constraint: There are temporal relationships between certain atomic tasks (e.g., "seeding" must follow "fertilizing"), expressed as a set of binary tuples C. prec ={(a, b) | a must precede b} means that any feasible sequence S must satisfy these partial order relations. Kinematic constraints: These are satisfied during atomic task generation by ensuring that the radius of curvature of the path is greater than the minimum turning radius of the agricultural machine.

[0034] The optimization algorithm engine submodule is used to search and calculate the atomic task execution sequence with the objective function of maximizing the overall efficiency of the job and the core constraint of the total energy, and output at least one candidate optimized job sequence. Taking a genetic algorithm as an example, it includes: Encoding and Population Initialization: One chromosome represents an atomic task sequence S. The initial population is generated randomly, and a sequence constraint filter is used to ensure the legitimacy of individuals.

[0035] Fitness assessment: First, examine the energy constraint; assign extremely low fitness to sequences that severely violate it. For feasible or slightly violated sequences, calculate the comprehensive benefit U(S), U(S) = w area ·A(S) / A total +w time ·T ideal / T(S); where A(S) is the effective work area completed by the sequence (geometric calculations of path coverage need to be considered); T(S) is the estimated total time of the sequence; w area and w timeThe weight set by the user, and w area +w time =1; A total and T ideal for normalization; A total refers to the total area of the target atomic task set by the user, which is the benchmark value of the work benefit; T ideal refers to the shortest theoretical time required to complete all atomic tasks under ideal working conditions without considering energy constraints and at the highest efficiency speed, which is the normalized benchmark of time benefit; work area benefit A(S) / A total represents the proportion of the sequence completion area to the total area; work time efficiency benefit T ideal / T(S) represents the ratio of ideal time T ideal to actual estimated time T(S) (T ideal is a constant, and the larger the value indicates that T(S) is smaller, reflecting that the work speed is faster); , where L i is the length (distance) of atomic task i, v i is the speed of executing the atomic task, L i / v i is the theoretical time of the atomic task, and T turn is the cumulative time sum of all non-traveling fixed actions such as turns in the sequence.

[0036] The sequence and budget output submodule is used to allocate an energy budget based on the estimated energy consumption for each atomic task in the selected optimized work sequence.

[0037] Genetic operation: adopt order crossover and exchange mutation and other operators specially designed for permutation coding to maintain the legality of offspring; Output: after the algorithm converges, output one or more sequences S opt with the highest fitness.

[0038] The sequence and budget output submodule allocates a budget opt for each atomic task i in S , where α is a budget buffer coefficient (such as 5%), and refers to the estimated energy consumption value of the i-th atomic task in the optimized sequence Sopt, which is calculated by calling the corresponding heterogeneous energy consumption model library submodule calculation formula (physical energy consumption model) according to its type (movement, work, or action).

[0039] The optimization algorithm engine sub-module is used to maximize the comprehensive benefit of the task as the objective function, and to search and calculate the atomic task execution sequence with the total energy constraint as the core constraint, that is, to take all atomic task arrangements that meet the agricultural order constraint as the solution space, and to use genetic algorithm for iterative heuristic search; The process is as follows: first, randomly generate N legal sequences as the initial population, then in each iteration of the generation, calculate the fitness of each sequence (take the energy constraint as the hard threshold, and take the weighted and normalized benefit U(S) of the task area and timeliness as the optimization target), and then select by roulette, and then use the order crossover and exchange mutation operator to generate a new population; This process is guided by the fitness function in multiple generations of evolution, continuously approaching sequences with higher comprehensive benefit and feasible energy, until convergence, and finally output one or more candidate optimized task sequences with the highest fitness.

[0040] In specific implementation, the real-time energy consumption monitoring and budget verification module includes a multi-source sensor fusion sub-module, an atomic task execution tracking sub-module, and an energy consumption verification and deviation calculation sub-module; The real-time energy consumption monitoring and budget verification module realizes the closed-loop connection of planning and execution, and is a "sensor" for discovering energy deviation and triggering re-planning.

[0041] The multi-source sensor fusion sub-module is used to collect battery electrical parameters, motor parameters, vehicle position and working mechanism state data in real time; The multi-source sensor fusion sub-module marks the data stream from different hardware (CAN bus, GNSS, IMU) with high-precision unified time stamps. The multi-source sensor fusion sub-module calculates the battery instantaneous power P bat (t) = V bat (t) × I bat (t), V bat (t) and I bat (t) are the battery instantaneous voltage and current, respectively, which are detected by the sensor in real time. The multi-source sensor fusion sub-module fuses GNSS position, IMU attitude and agricultural machine sensor signals to determine the accurate state of the vehicle (such as "straight driving", "turning", "agricultural machine has fallen").

[0042] The atomic task execution tracking submodule is used to determine the atomic task currently being executed according to the vehicle position and working mechanism state data; the submodule is implemented by a deterministic finite state machine, and the state transition of the machine is driven by multi-source data fusion events. The specific determination logic is: continuously receiving the "vehicle state" message fused with GNSS position, IMU heading and working mechanism sensor signals (such as hydraulic cylinder travel switch); matching the state with the pre-stored geometric path of the atomic task sequence. The state transition conditions are: 1) position trigger: when the vehicle center projection point enters the buffer zone of the atomic task planning path (such as path lateral offset <0.5m), and the included angle with the path tangent is less than a threshold, the "moving / working atomic task" is triggered; 2) signal trigger: for action atomic tasks (such as "lifting plow"), when the corresponding hydraulic pressure switch or position sensor signal is received, it is determined that the atomic task starts / ends; 3) distance / time trigger: for moving / working atomic tasks, the cumulative travel distance along the path is used to verify the progress in real time, and when the cumulative travel distance ≥ the atomic task length L, it is determined that the atomic task ends.

[0043] The energy consumption verification and deviation calculation submodule is used to accumulate the actual energy consumption of the current atomic task, and calculate the deviation value ΔE of the actual energy consumption from the corresponding energy budget. During the time period [t start , t end ] when the current atomic task j is executed, the actual energy consumption is obtained by integrating the instantaneous power: ; wherein, represents the actual energy consumption of the current atomic task j, and the unit is joule (J) or commonly used kilowatt-hour (kWh), which is the real value of energy consumption measured and accumulated by the system through sensors. P bat (t) represents the instantaneous output power of the battery at time t, and the unit is watt (W), which is obtained by multiplying the real-time collected battery voltage V bat (t) and current I bat (t), that is, P bat (t) = V bat (t) x I bat (t), which reflects the total power demand of the vehicle and the working mechanism at each moment. t start and t end represent the start time and end time of the current atomic task j, respectively, which are accurately determined by the state machine of the "atomic task execution tracking submodule". The integral represents the integration of the instantaneous power P bat (t) in the time interval [t start , t endThe physical meaning of this is to calculate the total energy output by the battery during the execution period of the atomic task. Through the integration of the real-time power of the battery during the execution period of the atomic task, the actual total energy consumed to complete the atomic task is accurately measured.

[0044] The relative deviation is calculated as follows: ; ΔE j represents the relative energy consumption deviation of atomic task j, which is a dimensionless ratio indicating the degree of deviation of actual consumption from the budget. represents the energy budget of atomic task j, which is pre-allocated by the "sequence and budget output submodule" in the planning stage, and is the planned value obtained by adding a certain management margin (a) to the estimated energy consumption based on the physical energy consumption model; the management margin (a) is based on the theoretical estimated energy consumption of the physical energy consumption model , a part of "safety pad" or "planned fault tolerance space" is actively added to absorb and respond to short-term random disturbances (such as slight road undulations, wind changes) and systematic modeling errors / simplifications (such as the calibration deviation of transmission efficiency η), so as to increase the budget of a single atomic task from "most likely value" to "more conservative guaranteed value". Its engineering role is reflected in two aspects: 1) improve the reliability of local atomic task completion: even if the actual energy consumption is slightly higher than the theoretical estimate, since there is a buffer of a, the atomic task can still be completed within the budget, avoiding frequent triggering of alarms due to minor fluctuations; 2) serve the global early warning mechanism: it sets a more stringent internal benchmark for subsequent real-time monitoring and deviation calculation. Only when the actual energy consumption exceeds this buffered budget, i.e., |ΔE j |>δ, it indicates that there is a significant and irregular overspending, which is likely to be caused by unexpected major disturbances (such as sudden hard soil), which enables the system to more accurately distinguish between "normal fluctuations" and "abnormal situations", ensuring that the re-planning mechanism is only triggered when it is really necessary, improving the judgment accuracy and stability of the entire system. Therefore, a is a key design parameter that connects "static planning" and "dynamic adaptation".

[0045] represents the absolute energy consumption deviation, and the positive value indicates overspending (actual budget is more than budget), and the negative value indicates surplus. This formula quantitatively evaluates whether the execution of the current atomic task deviates from the expected energy planning by calculating the relative deviation of actual energy consumption and planned budget. A large positive value indicates a risk of out-of-control energy consumption.

[0046] When |ΔE jWhen the absolute value of the deviation exceeds the threshold δ (e.g., δ=0.15), the atomic task is marked as having abnormal energy consumption. When the absolute value of the deviation exceeds the threshold δ, the system determines that the atomic task has "abnormal energy consumption." This indicates that the actual execution has significantly deviated from the planned expectations, possibly due to unforeseen field conditions (such as abnormally hard soil) or system failure. A clear and quantifiable alarm trigger line is set. Once triggered, a signal is sent to the "Dynamic Replanning Decision Module" to initiate a reassessment and optimization of subsequent atomic tasks, thereby ensuring that the constraints of the global energy budget are not ultimately breached.

[0047] In practical implementation, the dynamic replanning decision module includes an impact assessment submodule, a replanning triggering and strategy selection submodule, and a local sequence regeneration submodule. The dynamic replanning decision module endows the system with adaptability and robustness in the face of field uncertainties, which is the key to ensuring the feasibility of "atomic mission-level energy budget management" in practice.

[0048] The impact assessment submodule is used to assess the impact on the total energy budget of the remaining atomic tasks based on the energy consumption deviation of the current atomic task; the impact assessment submodule receives the energy consumption deviation ΔE of atomic task j. j Then, immediately reassess the global energy state: definition R represents the ratio of the remaining energy budget to the estimated demand for remaining atomic missions. It is a dimensionless ratio and is the core decision variable that triggers replanning. Budget for remaining available energy; E battery This represents the total available energy of the electric agricultural machinery's onboard battery at the current moment. This is a real-time value that varies with charging and discharging states. 'r' represents the global safety factor (0 < γ ≤ 1, e.g., 0.9), which indicates the maximum proportion of battery energy that is planned to be used. A portion of energy is reserved for emergency return trips or unmodeled losses, reflecting the system's safety. E battery •r represents the total available budget, which is the maximum energy limit that the system can utilize throughout the entire operating cycle. This represents the actual energy consumption of all completed atomic tasks. The sum of these values ​​represents the energy actually consumed by the battery. The instantaneous power is obtained by multiplying the voltage and current data collected in real time by the battery management system, and then numerically integrating this power value during the execution time of the atomic task. The sum of the baseline energy consumption for all unfinished atomic tasks. Based on the baseline for estimated energy consumption, the definition is similar to that above. , This represents the k-th incomplete atomic task. This represents the "expected" energy consumption most likely to be used to complete the atomic task; the baseline estimate is used here. but not the budget value allocated with the management margin a This is because which represents the “expected” energy consumption most likely to occur for completing the atomic task, is more objective and direct for evaluating future demand, the budget value is the value for control with the management margin a added on this basis. represents the estimated energy demand of the remaining atomic tasks; is the sum of the baseline estimated energy consumption of all unfinished atomic tasks. The formula as a whole represents that the estimated energy demand of the remaining atomic tasks is based on the initial model, the total prediction of the energy required for all unfinished atomic tasks, which represents the “expected expenditure” required for continuing execution as planned.

[0049] The re-planning trigger and strategy selection submodule is used to trigger re-planning when the ratio of the remaining total budget to the total estimated energy consumption of the remaining atomic tasks is lower than a preset threshold, and to select one of the parameter adjustment strategy, sequence adjustment strategy or insertion charging strategy. The trigger condition is R<β but R>0.8, and the parameter adjustment is selected; wherein β (usually set to a value slightly larger than 1, for example 1.1) represents the “energy safety margin” or “warning line” preset by the system to cope with future uncertainty, and the physical meaning of β is to quantitatively define the critical state in which the system considers that the remaining energy is starting to be tight and needs to intervene in advance. It contains two levels of buffer: 1) time buffer: reserves time and energy for re-planning calculation and new strategy execution; 2) risk buffer: for the further energy consumption fluctuations that may occur in the actual execution of the remaining atomic tasks (its prediction The main parameters of parameter adjustment include the working speed v and the working depth d. R < β means that there is a clear energy warning, but R > 0.8 indicates that the energy gap is still mild (the remaining budget is still greater than 80% of the estimated demand). At this time, the nature of the problem is that the trend is not good, but there is still sufficient time and energy for moderate and fine adjustment. The reasons for selecting parameter adjustment at this time are: 1) the cost is the smallest: parameter adjustment does not change the macro plan of the job (does not delete atomic tasks, does not change the order, does not interrupt the job), and the impact on the quality of the job and the overall job flow is the smallest, and the user's perception is the lightest; 2) the effect is reversible and fine: it is a "fine tuning" of the execution process, such as slightly reducing the speed, and if the subsequent situation improves, the original parameters can be restored. This way can accurately "fill the valley with the peak", and the required energy saving is exchanged for the smallest job quality cost; 3) it conforms to the operation common sense: when the fuel / electricity is slightly insufficient, the first reaction of the driver is also "to drive slowly to save fuel / electricity", rather than immediately deciding "not to go to a certain place" or "to find a gas station", and parameter adjustment simulates this most natural and first adopted optimization behavior. Why not choose other strategies at this time? The reasons include: 1) sequence adjustment (skip atomic tasks): this is a "structural adjustment" that will directly sacrifice job area and output, and is a large adjustment measure. When the energy gap is not serious (R > 0.8), the first choice should be to preserve the area and plan as much as possible, so skipping atomic tasks is an overreaction; 2) insert charging: this will interrupt the job and generate a lot of invalid empty driving (round trip to the charging station) time, which will seriously reduce the efficiency of the job. Only when the energy is severely deficient (R ≤ 0.5) and it is impossible to complete the atomic task through internal adjustment, is it worth paying such a large efficiency cost.

[0050] If 0.5 < R ≤ 0.8 and there are low-priority atomic tasks that can be skipped, sequence adjustment is selected; "sequence adjustment" refers to permanently removing one or more low-priority atomic tasks from the sequence of atomic tasks to be executed according to pre-set or dynamically calculated priorities, and then re-optimizing the execution order of the remaining atomic tasks, in order to maximize the completion of high-value jobs within the tight energy budget. The specific adjustment means is: first, automatically identify or mark according to user presets the low-priority atomic tasks that can be skipped according to atomic task attributes (such as the location of the field block to which they belong, the type of work); then, delete these atomic tasks from the current job sequence; then, call the genetic algorithm of the global sequence optimization module as the optimization algorithm engine, and quickly recalculate a new execution order that is optimal in terms of comprehensive benefits under the new energy constraints only with the remaining atomic tasks as the object; finally, output and update a new atomic task sequence that is shorter and more energy-efficient, and its corresponding revised energy budget.

[0051] If R≦0.5 or no atomic task can be skipped, and there is a charging station nearby, select to insert charging. Atomic task priority is preset by user or automatically determined according to atomic task type and location of the field block to which the atomic task belongs.

[0052] In summary, the parameter adjustment strategy mainly includes reducing the working speed or working depth of subsequent non-critical atomic tasks; the sequence adjustment strategy refers to skipping atomic tasks with lower priority; and the insertion charging strategy refers to planning a path to the nearest charging station and inserting it into the current working sequence.

[0053] The local sequence regeneration submodule is configured to call the global sequence optimization module to quickly re-optimize the remaining unexecuted atomic tasks according to the selected strategy, including: 1) parameter adjustment implementation: taking the speed v of all subsequent "working atomic tasks" as an optimized variable set V. A sub-optimization problem is constructed: under the premise that the total energy consumption after adjustment is less than or equal to , find v to make the total working time shortest or the working area largest. This is a continuous or discrete optimization problem, which can be quickly solved by gradient descent or enumeration method, and the details are as follows: taking the speed variable v of all subsequent "working atomic tasks" (forming a vector V) as the decision variable, the feasible region of each v is defined as [v min , v nominal ], where v nominal is the original set speed, and v min is the minimum allowed speed to ensure working quality (such as 70% of the rated speed); the optimization goal is to maximize the comprehensive benefit after adjustment, which can be in the form of minimizing the total working time or maximizing the effective working area A adj (V) (when the speed affects the coverage quality), where L i is the working length or distance of the i-th task, and the estimated time consumption of the i-th task at the adjusted speed v i . The core constraint is that the total estimated energy consumption after adjustment must be less than or equal to the current remaining energy budget , where is the new estimated energy consumption calculated by substituting v i into the physical energy consumption model of the task and considering the influence of speed on working resistance, and the solving method is: if the speed is regarded as a continuous variable and the target / constraint function is smooth, the gradient descent method can be used, and V （k+1） =V （k） -λ∇F(V （k）) to find the optimal solution. If the speed is considered as a finite number of discrete levels for the sake of controlling the discretization, the enumeration method is used to calculate the objective value and energy consumption under all speed combinations V, filter out the combinations violating the constraints, and then select the combination with the optimal objective value (shortest time or largest area) as the adjustment scheme. This sub-optimization problem is independent and small in size, which can be quickly solved by calling the numerical calculation kernel in the same optimization algorithm engine described above, and the output result is the optimal speed adjustment vector V ∗ , and accordingly update the instruction parameters of all subsequent related tasks. The iteration formula V （k+1） = V （k） - λ∇F(V （k） ) indicates that the speed vector is updated in the opposite direction of the gradient (i.e. the function descent direction) to gradually approach the optimal solution that satisfies the constraints: , where F is the Lagrange function, which includes the objective function and the energy consumption constraint expressed in the form of a penalty function; V （k） is the speed vector at the kth iteration, λ is the learning rate or step size, which is a positive number, controlling the adjustment amplitude of each iteration; F(V （k） ) is the Lagrange function or augmented objective function at the iteration point V （k） . It combines the original objective function (such as T adj (V)) and the penalty term of violating the constraints together, in the form of , where u is the penalty coefficient.

[0054] ∇F(V （k） ) represents the gradient (i.e. the first derivative vector) of the function F at the point V （k） , indicating the direction that makes F descend the fastest.

[0055] 2) Sequential adjustment implementation: remove the subset S remaining of atomic tasks with priority lower than the threshold P th from the remaining atomic task list S skip , to get a new list S new =S remaining -S skip , remove all elements belonging to the subset S remaining from the full set S skip , to get a new set S new . This is a set operation, not an arithmetic subtraction. Then send S new to the global sequence optimization module for reordering optimization (because after skipping, the optimal order between atomic tasks may change).

[0056] 3) Insert charging implementation: calculate the optimal path from the current position to the nearest charging station (as a new "mobile atomic task"), and the optimal connection point after charging from the charging station to return to the field to continue work. Insert the "going mobile", "charging", and "returning mobile" three new atomic tasks into the current sequence to form an expanded planning problem, and call the global optimization module to solve.

[0057] The core idea of the present application is to propose a new mode of "atomic task level energy budget management" to solve the technical problems of low reliability of atomic task completion and low energy utilization efficiency of electric agricultural machinery due to limited battery energy. Traditional solutions are mostly in real-time control at the bottom or use fixed mode, lacking global and adaptive ability. The present application realizes through the technical means of "decomposition-optimization-monitoring-replanning": 1) decompose macro work into three types of heterogeneous atomic tasks of moving, working, and acting and establish a physical energy consumption model; 2) globally optimize the atomic task execution sequence and allocate energy budget with energy as the constraint and comprehensive benefit (area, timeliness) as the target; 3) real-time monitor atomic task execution and energy consumption verification; 4) when the deviation threatens the overall atomic task, dynamically re-plan (adjust parameters, sequence or insert charging). These means constitute a closed-loop intelligent system of perception, planning, and execution. The technical effects achieved are: 1) maximize the effective work amount per charging from a global perspective, avoiding "mileage anxiety"; 2) through fine heterogeneous energy consumption perception and dynamic adjustment, improve the atomic task completion reliability and predictability of complex farmland work; 3) improve the intelligent level and practical value of electric agricultural machinery by upgrading energy management from passive response to active intelligent planning.

Claims

1. An automatic control system for electric agricultural machinery, characterized in that, It includes an atomic task decomposition and energy consumption modeling module, a global sequence optimization module, a real-time energy consumption monitoring and budget reconciliation module, a dynamic replanning decision-making module, and a human-computer interaction and execution interface module, with each module connected in sequence; The atomic task decomposition and energy consumption modeling module is used to decompose the macro-level farmland operation atomic task input by the user into multiple atomic tasks with different energy consumption characteristics, and to establish a basic energy consumption model for each type of atomic task. The global sequence optimization module is used to optimize the execution sequence of the atomic tasks under the total energy constraint with the goal of maximizing the overall efficiency of the operation, and to allocate an energy budget to each atomic task in the optimized job sequence. The real-time energy consumption monitoring and budget reconciliation module is used to track the currently executed atomic tasks and accumulate their actual energy consumption during agricultural machinery operations, and calculate the deviation between them and the allocated energy budget. The dynamic replanning decision module is used to trigger the re-optimization of the execution sequence or parameters of the remaining unexecuted atomic tasks when the deviation exceeds a threshold and the assessment will affect the completion of the overall atomic tasks, thereby generating updated energy budget and control instructions.

2. The automatic control system for electric agricultural machinery as described in claim 1, characterized in that, The atomic task decomposition and energy consumption modeling module includes: a map and atomic task parsing submodule, which is used to parse a digital map containing fields, obstacles and charging station locations, as well as user-specified task types and ranges; The atomic task generation submodule is used to decompose macroscopic operation atomic tasks into standardized atomic task sequences based on the parsing results. The atomic tasks include at least three types: movement atomic tasks, operation atomic tasks, and mechanism motion atomic tasks. The Heterogeneous Energy Consumption Model Library submodule is used to store the energy consumption calculation formulas and basic parameters corresponding to the three types of atomic tasks mentioned above.

3. The automatic control system for electric agricultural machinery as described in claim 2, characterized in that, For linear movement atomic tasks, the heterogeneous energy consumption model library submodule estimates the energy consumption E of the moving atomic task. move The calculation formula is: E move =(f roll ·m·g·cosθ+m·g·sinθ+1 / 2·ρ·C d ·A·v 2 )·L / h; Among them, f roll Where m is the rolling resistance coefficient, g is the mass of the agricultural machinery, θ is the slope angle, ρ is the air density, and C is the rolling resistance coefficient. d η is the drag coefficient, A is the frontal area, v is the set speed, L is the travel distance, and η is the transmission efficiency.

4. The automatic control system for electric agricultural machinery as described in claim 3, characterized in that, For field operation-type atomic tasks, the heterogeneous energy consumption model library submodule estimates the energy consumption E of the operation atomic task. work The calculation formula is: ； in, For the basic working resistance of agricultural machinery, H soil L is the soil hardness coefficient, k is the proportionality coefficient, and L is the soil hardness coefficient. work η represents the working distance and η represents the transmission efficiency.

5. The automatic control system for electric agricultural machinery as described in claim 4, characterized in that, For atomic tasks involving mechanism actions, the heterogeneous energy consumption model library submodule estimates the energy consumption E of the atomic task involving mechanism actions. action The calculation formula is: E action =P hydraulic t action ; Where P hydraulic τ is the rated power of the hydraulic system. action The duration of the action.

6. The automatic control system for electric agricultural machinery as described in claim 5, characterized in that, The global sequence optimization module includes a constraint management submodule, an optimization algorithm engine submodule, and a sequence and budget output submodule. The constraint management submodule is used to define and manage total energy constraints, atomic task sequence constraints, and agricultural machinery kinematic constraints. The optimization algorithm engine submodule is used to search and calculate the atomic task execution sequence with the objective function of maximizing the overall efficiency of the job and the core constraint of the total energy constraint, and output at least one candidate optimized job sequence. The sequence and budget output submodule is used to assign an energy budget based on the estimated energy consumption of each atomic task in the selected optimized job sequence.

7. The automatic control system for electric agricultural machinery as described in claim 6, characterized in that, The optimization algorithm engine submodule uses a genetic algorithm for heuristic search, and the overall job efficiency U(S) is expressed as: U(S)=w area ·A(S) / A total +w time ·T ideal / T(S); Where A(S) is the effective work area completed by the sequence (geometric calculations of path coverage need to be considered); T(S) is the estimated total time of the sequence; w area and w time The weights set for the user, and w area +w time =1; A total and T ideal Used for normalization; A total The total area of ​​the target atomic tasks set by the user is the benchmark value for operational efficiency; T ideal The shortest theoretical time required to complete all atomic tasks under ideal conditions of continuous operation at maximum efficiency, without considering energy constraints, serves as a normalized benchmark for time efficiency; operating area benefit A(S) / A total This represents the proportion of the area completed in the sequence to the total area; the time efficiency benefit T. ideal / T(S) represents the ideal time T. ideal The ratio of the actual estimated time T(s) to the actual estimated time (the larger the value, the faster the operation). L i v represents the length (distance) of atomic task i. i To determine the speed at which this atomic task can be performed, L i / v i This is the theoretical time taken for the atomic task, T. turn This is the total cumulative time taken for all non-fixed movement actions such as turning at the edge of the terrain in the sequence.

8. The automatic control system for electric agricultural machinery as described in claim 7, characterized in that, The real-time energy consumption monitoring and budget reimbursement module includes: a multi-source sensor fusion submodule, an atomic task execution tracking submodule, and an energy consumption reimbursement and deviation calculation submodule; The multi-source sensor fusion submodule is used to collect real-time data on battery electrical parameters, motor parameters, vehicle position, and operating mechanism status. The atomic task execution tracking submodule is used to determine the currently executing atomic task based on vehicle location and operating mechanism status data; The energy consumption verification and deviation calculation submodule is used to accumulate the actual energy consumption of the current atomic task and calculate its deviation value ∆E from the corresponding energy budget.

9. The automatic control system for electric agricultural machinery as described in claim 8, characterized in that, The dynamic replanning decision-making module includes an impact assessment submodule, a replanning triggering and strategy selection submodule, and a local sequence regeneration submodule. The impact assessment submodule is used to assess the impact on the total energy budget of the remaining atomic tasks based on the energy consumption deviation of the current atomic tasks. The replanning triggering and strategy selection submodule is used to trigger replanning when the ratio of the remaining total budget to the remaining total estimated energy consumption of atomic tasks is lower than a preset threshold, and select one of the parameter adjustment strategy, sequence adjustment strategy or insertion charging strategy. The local sequence regeneration submodule is used to call the global sequence optimization module to re-optimize the remaining unexecuted atomic tasks for the selected strategy.

10. The automatic control system for electric agricultural machinery as described in claim 9, characterized in that, The parameter adjustment strategy refers to reducing the job speed or job depth of subsequent non-critical atomic tasks; the sequence adjustment strategy refers to skipping low-priority job atomic tasks; and the insertion charging strategy refers to planning a path to the nearest charging station and inserting it into the current job sequence.

Images