A task allocation method and system based on hybrid game genetic algorithm

Abstract

This invention discloses a task allocation method and system based on a hybrid game genetic algorithm, belonging to the field of intelligent manufacturing. The method comprises the following steps: acquiring task allocation data based on the Industrial Internet of Things and edge computing technologies; constructing a collaborative optimization model for task allocation, path planning, and energy consumption control based on the task allocation data; the collaborative optimization model includes a three-dimensional optimization objective function and a constraint system; the constraint system includes: equipment health constraints, AGV dynamic motion constraints, process connection spatiotemporal constraints, and multi-agent game equilibrium constraints; and solving the collaborative optimization model using a hybrid game genetic algorithm to obtain the optimal solution for task allocation. By deeply coupling multi-agent games with dynamic evolutionary algorithms, the present invention overcomes the problem of split decision dimensions in traditional single-objective static scheduling.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent manufacturing technology, and in particular relates to a task allocation method and system based on a hybrid game genetic algorithm. Background Art

[0002] Currently, task allocation in textile workshops primarily relies on traditional scheduling algorithms (such as heuristic algorithms and genetic algorithms). However, these algorithms face the following key bottlenecks in multi-objective collaborative optimization scenarios: ① Local optimality trap: Traditional genetic algorithms employ fixed crossover / mutation probabilities, making them prone to falling into local optimal solutions. This makes it particularly difficult to balance various performance metrics, particularly when optimizing for multiple objectives (such as completion time, energy consumption, and idle rate). ② Imbalanced resource allocation: The separation of task allocation and equipment selection leads to uneven equipment loads; AGV path planning is independent of production plans, resulting in high idle rates and significant energy redundancy. ③ Insufficient dynamic adaptability: The lack of a real-time data feedback mechanism makes it incapable of responding to dynamic disturbances such as order insertion and equipment failures; rigid charging strategies lead to AGV interruptions due to insufficient power. ④ Ineffective multi-objective optimization: Traditional algorithms focus solely on a single objective (such as minimizing completion time), ignoring the coordinated optimization of energy consumption and logistics efficiency. Path planning also fails to consider the dynamic occupancy of warehouse resources, leading to warehouse conflicts.

[0003] The limitations of the existing technology essentially stem from the design paradigm of "single goal orientation" and "static decision-making". Therefore, the present invention provides a task allocation method based on a hybrid game genetic algorithm. Summary of the Invention

[0004] In order to solve the above technical problems, the present invention proposes a task allocation method and system based on a hybrid game genetic algorithm to solve the problems existing in the above prior art.

[0005] To achieve the above object, the present invention provides a task allocation method based on a hybrid game genetic algorithm, comprising:

[0006] Obtain task allocation data based on industrial Internet of Things and edge computing technology;

[0007] A collaborative optimization model of task allocation, path planning, and energy consumption control is constructed based on the task allocation data; the collaborative optimization model includes a three-dimensional optimization objective function and a constraint system; the constraint system includes: equipment health status constraints, AGV dynamic motion constraints, process connection spatiotemporal constraints, and multi-agent game equilibrium constraints;

[0008] The collaborative optimization model is solved based on a hybrid game genetic algorithm to obtain the optimal solution for task allocation.

[0009] Optionally, the task allocation data includes: equipment status data, AGV charge status data, warehouse topology and order dynamic data.

[0010] Optionally, the expression of the three-dimensional optimization objective function is:

[0011]

[0012]

[0013]

[0014] Where, represents the maximum process completion time, Indicates the completion time of the mth process, m represents the process number, and M represents the total number of processes. represents the total energy consumption, N represents the total number of devices, represents the power of the i-th device, represents the processing time of the i-th device, Indicates the total number of AGVs, represents the energy consumption during the operation of the jth AGV, represents the idle distance of the jth AGV, represents the total travel distance of the jth AGV, Indicates the AGV idle rate.

[0015] Optionally, the expression of the device health status constraint is:

[0016]

[0017] The expression of the AGV dynamic motion constraint is:

[0018]

[0019] Where, represents the health index of the i-th device, represents the path node conflict detection function, Indicates the indicator function, which takes 1 when in charging state. represents the health index of the i-th device at time t, represents the absolute value of the acceleration of the j-th AGV at time t, Indicates the status of AGV charging and running. represents the charge state of the j-th AGV at time t, k represents the k-th task, It represents the path node at time t when the j-th AGV performs the k-th task. It represents the path node at time t when the lth AGV performs the mth task. Indicates the difference from AGV number, Indicates the moment, Indicates the AGV number, represents the state of the j-th AGV at time t, Indicates the total number of charging piles.

[0020] Optionally, the expression of the process connection time and space constraint is:

[0021]

[0022] The expression of the multi-agent game equilibrium constraint is:

[0023]

[0024] Where, represents the start time of task k, Represents the arrival time of the precursor material p, which belongs to the set of precursor materials of task k , represents the set of precursor materials for task k, Indicates the total number of materials currently involved, Represents an indicator function. If the storage location of material x at time t is c, the value is 1, otherwise it is 0. represents the volume of material x, represents the volume of material x, represents the maximum capacity of storage location c, Indicates the end time of the mth process at the equipment end, It represents the time when AGV arrives at the mth process, Indicates the standard processing time of the current process. Represents the Shapley value of the device agent, represents the Shapley value of the AGV agent, Indicates the Shapley value of Energy Agent, represents the profit matrix of the three-party agents at time t, represents the mixed strategy space, represents the utility function of agent y, represents the Frobenius norm of the profit matrix at time t and time (t-1), represents the mixed strategy Nash equilibrium point, Indicates that agent y adopts a balanced strategy among other agents , adopt your own strategy The utility function value at .

[0025] Optionally, the process of solving the collaborative optimization model based on a hybrid game genetic algorithm includes:

[0026] The chromosome is divided into the equipment selection gene segment, the path planning gene segment, the energy consumption gene segment and the game strategy gene segment to obtain a four-segment dynamic coding structure.

[0027] Constructing a multi-agent game model, and dynamically adjusting the population corresponding to the four-segment dynamic coding structure to move toward the Pareto frontier based on updating the payoff matrix and Shapley value of the multi-agent game model;

[0028] The optimal solution for task allocation is obtained by adjusting the crossover operator and dynamic mutation operator in the process of the population moving towards the Pareto front through an adaptive crossover and mutation mechanism.

[0029] Optionally, the calculation expression of the Shapley value is:

[0030]

[0031] Where, represents the agent currently calculating the Shapley value, represents the Shapley value corresponding to agent a, represents the set of three intelligent agents {MA, AGV, EA}, express Any subset of Represents the marginal contribution of agent a to the alliance S.

[0032] Optionally, the adaptive crossover-mutation mechanism includes Pareto frontier detection crossover and dynamic mutation.

[0033] Optionally, a three-level mutation intensity control adaptive crossover mutation mechanism is adopted;

[0034] Among them, the three levels of variation intensity include: elite layer, middle layer and worst layer.

[0035] The present invention also provides a task allocation system based on a hybrid game genetic algorithm, which is used to implement the task allocation method. The system includes:

[0036] Multi-source data perception layer, used to obtain task allocation data based on industrial Internet of Things and edge computing technologies;

[0037] A collaborative optimization model layer, used to build a collaborative optimization model of task allocation, path planning, and energy consumption control based on the task allocation data;

[0038] Hybrid game genetic algorithm engine, used to integrate Nash bargaining mechanism and improved genetic algorithm, design dynamic chromosome encoding strategy, Pareto frontier detection operator, and multi-agent payoff matrix;

[0039] The real-time feedback control layer is used to simulate and verify the scheduling plan on the digital twin platform, and to issue dynamic adjustment instructions through the edge computing nodes.

[0040] Compared with the existing technology, the present invention has the following advantages and technical effects: the present invention constructs a three-dimensional collaborative optimization model of task allocation-path planning-energy consumption control. Through the deep coupling of multi-agent game and dynamic evolutionary algorithm, it breaks through the decision-making dimension separation problem of traditional single-objective static scheduling, integrates the objective function, and defines a three-dimensional optimization objective function to incorporate equipment processing time, AGV transportation energy consumption and empty driving rate into the same optimization system. Multi-objective collaborative optimization is achieved through dynamic weights, and a dynamic correlation model of energy consumption and efficiency is established. Through the collaborative optimization of acceleration constraints and charging strategies, the energy consumption redundancy of AGVs is reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The accompanying drawings, which constitute part of this application, are intended to provide a further understanding of this application. The exemplary embodiments and descriptions of this application are intended to explain this application and do not constitute an improper limitation on this application. In the accompanying drawings:

[0042] Figure 1 This is a system framework diagram of an embodiment of the present invention;

[0043] Figure 2 This is a flow chart of the algorithm mechanism of an embodiment of the present invention;

[0044] Figure 3 This is a flowchart of an adaptive crossover mutation process according to an embodiment of the present invention;

[0045] Figure 4 Feedback on the gaming strategy of the embodiment of the present invention;

[0046] Figure 5 This is a diagram showing the energy consumption optimization results of an embodiment of the present invention;

[0047] Figure 6 This is the change in the AGV idle rate according to the embodiment of the present invention. DETAILED DESCRIPTION

[0048] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in this application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0049] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0050] Example 1.

[0051] The limitations of existing technologies stem fundamentally from a single-goal-oriented and static decision-making design paradigm. Specifically, they are manifested in the following: 1. Lack of multi-goal coordination: Independent optimization of equipment scheduling and logistics transportation results in increased AGV waiting times and a high proportion of idling energy consumption. A lack of coupled optimization of charging decisions and transportation tasks results in a significant proportion of AGV non-operating energy consumption. 2. Weak dynamic constraint handling capabilities: Traditional algorithms rely on offline parameter tuning and are unable to adapt to real-time changes in the shop floor (such as fluctuations in equipment status and order priorities). Path planning utilizes a static A* algorithm, which cannot avoid sudden congestion and results in a high rate of path conflicts. 3. Imperfect agent collaboration mechanisms: Lack of effective negotiation among multiple agents (such as machine agents and AGV agents) leads to resource conflicts. Nash equilibrium solutions remain limited to theoretical models and lack closed-loop optimization with genetic algorithms, resulting in poor strategy stability. 4. Inherent limitations of the evolutionary operators of traditional genetic algorithms: Rigid crossover / mutation strategies: Fixed probability parameters lead to an imbalance between exploration and exploitation, making it easy to fall into local optima in the early stages, and a loss of population diversity in the later stages prevents the system from escaping suboptimal solutions. A single elite retention strategy: This relies solely on an elite selection mechanism based on fitness ranking, ignoring the Pareto dominance relationship between multiple objectives and resulting in an uneven distribution of solutions. Environmental response lag: The iteration process fails to incorporate real-time operating data (such as AGV battery life and equipment fault signals), resulting in significant deviations between offline optimization results and the actual scenario. ⑤ The disconnected optimization of task allocation, equipment scheduling, and logistics transportation leads to systemic losses: Spatial-temporal resource conflicts: AGV path planning is not synchronized with machine processing timing, resulting in a high proportion of AGV waiting time at the machine and increased machine idleness. Energy synergy blind spots: The lack of a correlation model between energy consumption and efficiency leads to excessive AGV acceleration and deceleration frequency, resulting in excessive motor energy redundancy. Charging scheduling mismatch: The charging station allocation strategy is decoupled from transportation tasks, resulting in increased task delays due to charging interruptions. Existing technologies face a fundamental bottleneck in their ability to respond to emergencies, namely, the failure of collaborative optimization within a distributed decision-making framework. A negotiation mechanism is missing: The machine agent and AGV agent adopt a simple priority preemption strategy, resulting in a high proportion of resource conflict resolution time. Game equilibrium distortion: Traditional Nash equilibrium solutions rely on a static payoff matrix and do not consider the state transition probability of the workshop, resulting in large errors in strategy stability.

[0052] In the context of intelligent upgrading in the textile industry, traditional single-objective static scheduling methods for complex production scenarios involving multiple coupled processes (spinning, weaving, and dyeing), heterogeneous equipment clusters (over 500 processing units), and large-scale AGV scheduling (over 100 transport units) suffer from systemic flaws such as fragmented decision-making dimensions (independent optimization of processes, equipment, and logistics), delayed dynamic response, and extensive energy efficiency management (a high proportion of inefficient energy consumption). The innovation of this invention lies in the construction of a three-dimensional collaborative optimization model for task allocation, path planning, and energy consumption management. By designing a chromosome dynamic encoding strategy guided by a Nash bargaining mechanism (the encoding dimensions include equipment operating conditions, AGV charge status, and warehouse topology information), an adaptive crossover mutation operator based on Pareto front detection, and a population evolution direction correction mechanism driven by a multi-agent payoff matrix, this method achieves multi-objective Pareto optimization of process connection efficiency, logistics turnover rate, and output per unit energy consumption. This method particularly overcomes the real-time decision-making bottleneck of traditional algorithms when operating on a scale of over 200 concurrent tasks.

[0053] The innovative features of the present invention are as follows: (1) Multi-agent strategy gene fusion coding: breaking through the limitation of traditional chromosomes that only map physical parameters, the game strategy is embedded in the gene structure.

[0054] Four-dimensional dynamic gene segment: Dynamic weight coding is integrated into the equipment health gene segment to trigger weight reconstruction of high-precision processes; the path planning segment adopts spatiotemporal conflict hot zone coding, dynamically generates congestion attenuation coefficients based on RFID grid occupancy, and introduces the exclusive-OR logic of charge state and charging pile occupancy to achieve gene-level response of conflict prediction and path avoidance; the energy consumption segment encodes energy efficiency parameters as field strength coupling factors through joint modeling of acceleration energy consumption field intensity and equipment health loss, driving energy consumption balance optimization of path and equipment selection; the game strategy segment adopts Shapley value-Nash equilibrium dual-mode coding, records the contribution ratio of intelligent agents and equilibrium convergence signs at the gene level, and adjusts the expression intensity of the strategy anchor area through real-time feedback from edge nodes, so that the chromosome has adaptive multi-objective game navigation capabilities.

[0055] (2) Hierarchical Perception Evolutionary Controller: This breaks through the static parameter limitations of traditional evolutionary operators and proposes a dynamic perception and intelligent control mechanism based on the Pareto frontier hierarchy. Through the triple innovation of hierarchical threshold decision-making, collaborative probability attenuation, and iterative adaptation, it realizes evolutionary control of multi-objective optimization.

[0056] ① Dynamic Threshold Decision-Making Mechanism: Frontier-Level Intelligent Perception: Real-time monitoring of the population's distribution density in the target space, dynamically adjusting the cross-trigger threshold based on the total number of frontier levels to prevent low-quality levels from interfering with high-quality solutions. ② Hierarchical Control Strategy: Dividing the population into a global exploration zone, a transition optimization zone, and an elite protection zone, with differentiated evolutionary rules designed for each zone. ③ Iterative Feedback-Compensation Mechanism: Dynamically adjusts threshold parameters and decay rates. When population diversity falls below a critical threshold, targeted mutations are implemented on individuals in the transition zone to prevent evolutionary stagnation.

[0057] (3) Game-evolution dynamic collaborative engine: Build a deep coupling mechanism between multi-agent game and genetic algorithm, and realize real-time evolutionary navigation through dynamic benefit feedback and strategy anchoring technology.

[0058] Dynamic profit matrix construction: Generate a profit matrix in real time based on equipment utilization, AGV effective transportation rate and energy efficiency index to quantify the collaborative contribution of intelligent agents; Shapley value gradient optimization: Dynamically adjust the objective function weight through alliance marginal contribution calculation to guide the population to move towards the Pareto frontier; Nash equilibrium convergence mark embedding: Use equilibrium error as a gene flag to trigger the adjustment of the strategy anchor area of ​​the edge node.

[0059] (4) Three-dimensional collaborative optimization model: A three-dimensional collaborative optimization model of task allocation, path planning and energy consumption control was constructed. Through the deep coupling of multi-agent game and dynamic evolutionary algorithm, the decision-making dimension separation problem of traditional single-objective static scheduling was broken through. The objective function was integrated and the three-dimensional optimization objective function was defined to incorporate equipment processing time, AGV transportation energy consumption and empty driving rate into the same optimization system. Multi-objective collaborative optimization was achieved through dynamic weights, and a dynamic correlation model of energy consumption and efficiency was established. Through the collaborative optimization of acceleration constraints and charging strategies, the energy consumption redundancy of AGVs was reduced.

[0060] like Figure 1 As shown in the figure, the present invention constructs a three-dimensional collaborative optimization framework based on a hybrid game genetic algorithm, and realizes the collaborative optimization of task allocation, path planning and energy consumption control in intelligent manufacturing workshops through the deep coupling of multi-agent game and dynamic evolution algorithm.

[0061] This embodiment provides a task allocation system based on a hybrid game genetic algorithm. The system adopts a four-layer architecture and achieves dynamic optimization through closed-loop control driven by data and intelligent decision-making: 1. Multi-source data perception layer: Integrates the Industrial Internet of Things (IoT) and edge computing technologies to collect three key data types in real time: equipment status (vibration / temperature sensors), AGV charge status (BMS data), warehouse topology (RFID positioning), and order dynamics (MES system data). 2. Collaborative optimization model layer: Constructs a three-dimensional objective function for task allocation, path planning, and energy consumption control, defining constraints for process connection, AGV charging time windows, and equipment load balancing. 3. Hybrid game genetic algorithm engine: Integrates the Nash bargaining mechanism with an improved genetic algorithm to design a dynamic chromosome encoding strategy, a Pareto frontier detection operator, and a multi-agent payoff matrix. 4. Real-time feedback control layer: Implements simulation verification of scheduling solutions based on a digital twin platform and issues dynamic adjustment instructions through edge computing nodes.

[0062] As a specific implementation method of this embodiment, this embodiment also provides a task allocation method based on a hybrid game genetic algorithm, which specifically includes the following steps: obtaining task allocation data based on industrial Internet of Things and edge computing technology; constructing a collaborative optimization model of task allocation-path planning-energy consumption control based on the task allocation data; the collaborative optimization model includes a three-dimensional optimization objective function and a constraint system; the constraint system includes: equipment health status constraints, AGV dynamic motion constraints, process connection time and space constraints and multi-agent game equilibrium constraints; solving the collaborative optimization model based on a hybrid game genetic algorithm to obtain the optimal solution for task allocation.

[0063] Multi-source data perception layer: This layer integrates three types of data through IIoT edge nodes. For equipment operating condition data, sensor types include piezoelectric vibration sensors (range 0-20g) and infrared temperature sensors (range 0-150°C).

[0064] Key parameters are as follows: ① Vibration RMS (Raw Mean Square) with a sampling rate of 10kHz to detect abnormal machining vibration; ② Bearing temperature with an accuracy of ±0.5°C to monitor overheating risks; ③ Extracting the Equipment Health Index (HI): .

[0065] AGV dynamic data: SOC monitoring: accuracy ±1%, sampling period 1 second; Indicates the RMS value of the device's current operating parameters (such as vibration amplitude, current, etc.), reflecting the device's real-time operating status; Indicates the RMS threshold of the corresponding parameter when the device is operating normally, which serves as the benchmark value for judging whether the device is normal. Indicates the current cumulative running time of the device, and records the running time since the device was put into use; Indicates the maximum cumulative operating time (or cycle) designed for the equipment, representing the theoretical maximum effective operating time of the equipment.

[0066] Real-time location: UWB positioning (accuracy ±10cm); Motion status: speed (0-2m / s), acceleration (±0.3m / s²); Charging strategy model: .

[0067] Indicates the charging time of the AGV, which is used to measure the time required to charge the AGV; Indicates the current state of charge of the AGV (State of Charge), which indicates the proportion of remaining battery power; Indicates the capacity of the AGV battery, that is, the total amount of electricity that the battery can store; Indicates the power of the AGV when charging, reflecting the amount of electricity charged into the battery per unit time.

[0068] Order and warehouse data: Obtain order priority (calculated based on delivery urgency and customer level) and real-time inventory coordinates (RFID grid resolution 0.5m×0.5m) from the MES system.

[0069] Collaborative optimization model layer: multi-objective modeling, in which a three-dimensional optimization objective function is defined.

[0070]

[0071]

[0072]

[0073] Where, represents the maximum process completion time, Indicates the completion time of the mth process, m represents the process number, and M represents the total number of processes. represents the total energy consumption, N represents the total number of devices, represents the power of the i-th device, represents the processing time of the i-th device, Indicates the total number of AGVs, represents the energy consumption during the operation of the jth AGV, represents the idle distance of the jth AGV, represents the total driving distance of the jth AGV (including empty driving and loaded driving), Indicates the AGV idle rate.

[0074] The constraint system includes: 1. Equipment health status constraint (based on HI index).

[0075] .

[0076] Constraint ①: Ensure that high-precision processing equipment (such as cotton yarn counts above 60) is in excellent condition; Constraint ②: Model the synergistic effects of equipment group degradation to prevent cluster failures; Constraint ③: Detect abnormal degradation (such as sudden bearing wear) through a sliding time window.

[0077] Where, represents the health index of the i-th device, represents the path node conflict detection function, Indicates the indicator function, which takes 1 when in charging state. represents the health index of the i-th device at time t, represents the absolute value of the acceleration of the j-th AGV at time t, Indicates the status of the AGV (such as charging, running), represents the charge state of the jth AGV at time t. In the acceleration limit formula, it ensures that the AGV moves more flexibly when the power is sufficient, and limits the acceleration when the power is low to avoid equipment failure or task interruption due to insufficient power. k represents the kth task, It represents the path node at time t when the j-th AGV performs the k-th task. It represents the path node at time t when the lth AGV performs the mth task. Indicates the AGV number (different from the AGV ID of j), Indicates the moment (time parameter), represents the AGV number (the jth AGV), represents the state of the j-th AGV at time t, Indicates the total number of charging piles.

[0078] 2. AGV dynamic motion constraints (integration of BMS and UWB data).

[0079] .

[0080] Represents the path node conflict detection function (grid resolution 0.5m); Indicates the total number of charging piles (dynamically dispatchable number); Indicates the indicator function, which is 1 when in charging state; is the acceleration limit; To avoid path conflicts; Occupied by charging piles.

[0081] 3. Time and space constraints for process connection (based on RFID warehouse topology).

[0082] .

[0083] represents the set of precursor materials for task k, represents the maximum capacity of storage location c, Indicates the standard processing time of the current process. represents the profit matrix of the three-party agents at time t, represents the mixed strategy space, represents the utility function of agent y, represents the start time of task k, Represents the arrival time of the precursor material p, which belongs to the set of precursor materials of task k , Indicates the total number of materials currently involved, represents the indicator function. If the storage location of material x at time t is c, the value is 1, otherwise it is 0; represents the volume of material x, Indicates the end time of the mth process at the equipment end, Indicates the time when the AGV arrives at the mth process. For material completeness; The storage capacity is limited; This is the AGV connection sequence.

[0084] 4. Equilibrium constraints for multi-agent games.

[0085] .

[0086] Shapley value of the machine agent; represents the Shapley value of the AGV agent; Indicates the Shapley value of the Energy Agent. It represents the Frobenius norm of the return matrix at time t and time (t-1), and is used to measure the magnitude of the matrix change. represents the mixed strategy Nash equilibrium point; Indicates that agent y adopts a balanced strategy among other agents , adopt your own strategy The utility function value at . Assign Shapley values; is the stability of the payoff matrix; The existence of Nash equilibrium.

[0087] Hybrid game genetic algorithm engine: Hybrid game genetic algorithm process is as follows Figure 2Dynamic chromosome encoding strategy: The chromosome adopts a four-segment dynamic encoding structure, and uses multi-dimensional gene mapping to map the real-time status and game strategy of the workshop. The specific expression is: .

[0088] 1. Equipment selection gene segment ( ): integer encoding represents the device ID + health status weight, each gene bit is defined as: .

[0089] 2. Path planning gene segment ( ): Binary encoding based on the spatiotemporal topology matrix, dynamically integrating warehouse occupancy information: .

[0090] in, ,β is updated through game theory, Indicates the logical and weighted integration of the path information of each task to generate a binary gene segment. k is a loop variable, which represents the sequence number of the currently processed task. It traverses each task from 1 to K. It is the total number of tasks and serves as the upper limit of k, indicating that there are K tasks involved in the logical and weighted integration. Represents the indicator function, if path j belongs to the set of feasible paths for task k , the function value is 1; if it is not feasible, the function value is 0, which is used to screen the effective path of task k. It reflects the congestion level of path j at time t, and the weight will be dynamically adjusted according to the real-time congestion status of the path.

[0091] 3. Energy consumption gene segment ( ). Floating point encoding represents energy efficiency decision parameters: .

[0092] 4. Game strategy gene segment ( ). Hybrid coding stores the ratio of multi-agent Shapley values ​​and the Nash equilibrium convergence sign: ,in, .

[0093] Fitness function: .

[0094] in, , It represents the Euclidean distance between the strategy vector p of the kth generation and the k-1th generation, and measures the magnitude of the change in strategy between the two generations. The larger the distance, the more significant the strategy change; the smaller the distance, the more stable the strategy. Through constant Normalize the distance so that the value is within a reasonable range. Affecting multiple target items: .

[0095] Adaptive crossover mutation mechanism: such as Figure 3 ① Pareto frontier detection intersection includes frontier level division: the population is divided into frontier levels according to non-dominated sorting (Rank 1 is the optimal frontier), and the threshold is adaptively adjusted according to the total number of frontier levels of the population: .

[0096] Dynamic adjustment of crossover probability: .

[0097] Gene fragment exchange rules: Through dynamic threshold adjustment (setting the threshold based on the total number of frontier levels of the population), nonlinear exponential decay strategy (maintaining a high crossover probability of 0.8 when ΔRank ≥ threshold to promote global search, and rapidly reducing the crossover probability by 0.7e^{-0.1ΔRank} when the difference is moderate to reduce the disturbance to the high-quality solutions of adjacent levels), and enhanced same-level protection (when ΔRank = 0, the crossover probability is reduced to 0.5 to avoid destroying the elite solution), a balance between exploration and development is achieved, ensuring global search capabilities while enhancing local development stability. It is particularly suitable for industrial optimization scenarios such as multi-objective workshop scheduling that require a balance between diversity and convergence speed.

[0098] ② Dynamic mutation includes the adaptive mutation formula based on the Pareto frontier. The adaptive mutation formula is: .

[0099] Adaptive rules for variable step length:

[0100]

[0101]

[0102] At the beginning of the iteration (t→0): β≈0.3 provides strong perturbation, α≈0 maintains a large step size; at the end of the iteration (t→T): β≈0 weakens the perturbation, α≈0.1 accelerates the step size decay; where: represents the basic mutation probability; Represents an individual Frontier level; : the total number of current population frontier levels; Indicates the maximum variable step length.

[0103] The three-level mutation intensity control includes the elite layer (Rank 1): the probability of mutation using attenuation protection decreases linearly with iteration, and the step attenuation rate is doubled; the middle layer (Rank 2~N-1): the standard adaptive mutation maintains the Rank-sensitive characteristics of the original formula; the worst layer (Rank N): the lower limit of the enhanced perturbation mutation probability is increased to 2 times the base value (upper limit 0.8), and the step size increases the dynamic enhancement coefficient β(t).

[0104] Multi-agent game optimization process: Construct a three-party game model of Machine Agent, AGV Agent, and Energy Agent, as shown in Table 1.

[0105] Table 1 Three-party game model.

[0106] Agent Type Revenue Function Parameter Description Machine Agent #timg# Equipment utilization AGV Agent #timg# Effective transportation rate Energy Agent #timg# Energy Efficiency Index

[0107] 1. Profit Matrix Construction: .

[0108] 2. Shapley value calculation optimization: .

[0109] Represents the agent currently calculating the Shapley value (such as Machine Agent, AGV Agent, EnergyAgent); represents the set of three intelligent agents {MA, AGV, EA}; express Any subset (coalition) of , representing a combination of participants that does not contain agent a, It is used to calculate the marginal contribution of a single agent to the alliance. The Shapley value of agent a needs to consider all subsets S that do not contain a, that is, S is Any subset of these subsets S represents the coalition without a participation, Represents the set of all remaining agents after removing agent a from set A. It represents the marginal contribution of agent a to the alliance S, that is, the increase in revenue after a joins.

[0110] Define the coalition value function For the Alliance The total benefit, based on the real-time benefit matrix calculate: . Through factorial terms Balance the contributions of alliances of different sizes, S=100.

[0111] 3. Nash equilibrium solution: .

[0112] The Shapley value is combined with the objective function gradient to ensure that the strategy moves toward the Pareto frontier. As the rate of change of the solution decays, convergence is accelerated. Termination conditions: and After each iteration, according to the current strategy Update the payoff matrix: , game strategy feedback such as Figure 4 shown.

[0113] Based on the above technical solution, the present invention has the following beneficial effects: Systematic improvement of multi-dimensional collaborative optimization capabilities: By integrating game equilibrium and dynamic evolution mechanisms, the synergy between task allocation, path planning, and energy consumption control is significantly enhanced. The coupled optimization of equipment utilization and logistics turnover efficiency effectively reduces idle resources. At the same time, through dynamic energy efficiency modeling, energy allocation is optimized, achieving efficient global resource scheduling in complex workshop scenarios, significantly reducing systematic losses in the production process.

[0114] Enhanced robustness and real-time performance in dynamic environments: A real-time data-driven game strategy update mechanism enables rapid response to emergencies such as order changes and equipment failures. By combining edge computing with digital twin technology, optimization plans can be dynamically adjusted based on shop floor status, reducing decision latency and ensuring scheduling stability and execution reliability in complex disturbance scenarios.

[0115] Breakthroughs in algorithm convergence efficiency and solution quality: Through an adaptive crossover mutation operator and a game equilibrium guidance mechanism, algorithm convergence is accelerated while maintaining population diversity. Pareto frontier detection technology effectively balances trade-offs between multiple objectives, generating a uniformly distributed set of high-quality solutions and significantly improving the overall performance of scheduling solutions.

[0116] Optimization of resource allocation fairness and energy efficiency coordination: Based on the contribution quantification mechanism of Shapley value, it avoids the resource grabbing or "free-riding" problem in traditional scheduling and realizes fair coordination between equipment, AGV and energy. Through the dynamic correlation modeling of energy consumption and efficiency, it optimizes the energy allocation of equipment processing and logistics transportation, significantly reducing ineffective energy consumption while ensuring production efficiency. Figure 5-Figure 6 shown.

[0117] Expanded adaptability in complex constraint scenarios: This algorithm leverages spatiotemporal topology encoding and dynamic gene mapping technologies to address equipment health constraints, avoid storage conflicts, and address process connection timing requirements, enhancing the algorithm's ability to schedule large-scale heterogeneous resources. A strategy negotiation mechanism within a hybrid game framework effectively resolves resource conflicts among multiple agents, improving the feasibility of solutions in complex constraint scenarios.

[0118] The above are merely preferred embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A task allocation method based on a hybrid game genetic algorithm, characterized in that: The following steps are involved: Obtain task allocation data based on industrial Internet of Things and edge computing technology; Building a collaborative optimization model of task allocation, path planning, and energy consumption control based on the task allocation data; The collaborative optimization model includes a three-dimensional optimization objective function and a constraint system; the constraint system includes: equipment health status constraints, AGV dynamic motion constraints, process connection time and space constraints, and multi-agent game equilibrium constraints; The expression of the three-dimensional optimization objective function is: Where, represents the maximum process completion time, Indicates the completion time of the mth process, m represents the process number, and M represents the total number of processes. represents the total energy consumption, N represents the total number of devices, represents the power of the i-th device, represents the processing time of the i-th device, Indicates the total number of AGVs, represents the energy consumption during the operation of the jth AGV, represents the idle distance of the jth AGV, represents the total travel distance of the jth AGV, Indicates the AGV empty driving rate; The expression of the equipment health status constraint is: The expression of the AGV dynamic motion constraint is: Where, represents the health index of the i-th device, represents the path node conflict detection function, Indicates the indicator function, which takes 1 when in charging state. represents the health index of the i-th device at time t, represents the absolute value of the acceleration of the j-th AGV at time t, Indicates the status of AGV charging and running. represents the charge state of the j-th AGV at time t, k represents the k-th task, It represents the path node at time t when the j-th AGV performs the k-th task. It represents the path node at time t when the lth AGV performs the mth task. Indicates the difference from AGV number, Indicates the moment, Indicates the AGV number, represents the state of the j-th AGV at time t, represents the total number of charging piles, and N represents the total number of devices; The expression of the time and space constraints of the process connection is: The expression of the multi-agent game equilibrium constraint is: Where, represents the start time of task k, Represents the arrival time of the precursor material p, which belongs to the set of precursor materials of task k , represents the set of precursor materials for task k, Indicates the total number of materials currently involved, Represents an indicator function. If the storage location of material x at time t is c, the value is 1, otherwise it is 0. represents the volume of material x, represents the maximum capacity of storage location c, Indicates the end time of the mth process at the equipment end, It represents the time when AGV arrives at the mth process, Indicates the standard processing time of the current process. represents the Shapley value of the device agent, represents the Shapley value of the AGV agent, Indicates the Shapley value of Energy Agent, represents the profit matrix of the three-party agents at time t, represents the mixed strategy space, represents the utility function of agent y, represents the Frobenius norm of the profit matrix at time t and time (t-1), represents the mixed strategy Nash equilibrium point, Indicates that agent y adopts a balanced strategy among other agents , adopt your own strategy The utility function value when ; Solving the collaborative optimization model based on a hybrid game genetic algorithm to obtain an optimal solution for task allocation; The process of solving the collaborative optimization model based on the hybrid game genetic algorithm includes: The chromosome is divided into the equipment selection gene segment, the path planning gene segment, the energy consumption gene segment and the game strategy gene segment to obtain a four-segment dynamic coding structure. Constructing a multi-agent game model, and dynamically adjusting the population corresponding to the four-segment dynamic coding structure to move toward the Pareto frontier based on updating the payoff matrix and Shapley value of the multi-agent game model; The adaptive crossover and mutation mechanism is used to adjust the crossover operator and dynamic mutation operator in the process of the population moving towards the Pareto frontier to obtain the optimal solution for task allocation. The calculation expression of the Shapley value is: Where, represents the agent currently calculating the Shapley value, represents the Shapley value corresponding to agent a, represents the set of three intelligent agents {MA, AGV, EA}, express Any subset of Represents the marginal contribution of agent a to the alliance S.

2. The task allocation method based on hybrid game genetic algorithm according to claim 1, characterized in that: The task allocation data includes: equipment status data, AGV charge status data, warehouse topology and order dynamic data.

3. The task allocation method based on hybrid game genetic algorithm according to claim 1, characterized in that: The adaptive crossover and mutation mechanism includes Pareto frontier detection crossover and dynamic mutation.

4. The task allocation method based on hybrid game genetic algorithm according to claim 3, characterized in that: Adopting three-level mutation intensity control to adaptive cross-mutation mechanism; Among them, the three levels of variation intensity include: elite layer, middle layer and worst layer.

5. A task allocation system based on a hybrid game genetic algorithm, characterized in that: For implementing the task allocation method according to claim 1, the system comprises: Multi-source data perception layer, used to obtain task allocation data based on industrial Internet of Things and edge computing technologies; A collaborative optimization model layer, used to build a collaborative optimization model of task allocation, path planning, and energy consumption control based on the task allocation data; Hybrid game genetic algorithm engine, used to integrate Nash bargaining mechanism and improved genetic algorithm, design dynamic chromosome encoding strategy, Pareto frontier detection operator, and multi-agent payoff matrix; The real-time feedback control layer is used to simulate and verify the scheduling plan on the digital twin platform, and to issue dynamic adjustment instructions through the edge computing nodes.

Images