CN115641049A - Multi-station multi-material task allocation and path planning method based on genetic algorithm - Google Patents

Abstract

The present invention provides a multi-station and multi-material task distribution and path planning method based on genetic algorithm, comprising steps: S1: a preprocessing step for material distribution; the S1 step further includes steps: S11: a modular design step; S12: Material clustering step; S13: Key material screening step; S14: Considering the demand of the station for batch distribution step; S2: Establishing a mathematical model with the shortest completion time as the optimization goal; S3: According to the constraint conditions, using the genetic algorithm to solve the mathematical model. The multi-station multi-material task distribution and path planning method based on genetic algorithm of the present invention aims at the problems of long-time and cumbersome distribution of multi-station and multi-material distribution in discrete assembly workshops, which restricts the completion of assembly tasks, etc., allowing multiple The station gets the correct type of material within the correct time window, and on the basis of completing the material distribution, the material distribution time is minimized as much as possible to improve the efficiency of the workshop.

Description

Multi-station multi-material task assignment and path planning method based on genetic algorithm

technical field

The present invention relates to the technical field of optimization methods for workshop material delivery routes, in particular to a multi-station and multi-material task assignment and route planning method based on genetic algorithms.

Background technique

(1) Patent document CN113344487A proposes an assembly line material distribution method and system based on a line-side integrated supermarket. The main technical solution is to use an improved harmony search algorithm based on the objective function of minimizing the cost of hiring workers and feeding costs , to determine the material distribution scheduling plan, the material distribution scheduling plan includes the responsible station of the material delivery worker, the time of each replenishment behavior, the station and the amount of material distributed for each station. In this way, the employment cost of material delivery workers and the cost of material delivery are minimized, and at the same time, the material distribution tasks are efficiently and reliably completed. (2) The patent document CN113947310A proposes a method based on the improved ant colony algorithm workshop material distribution path optimization method, the main technical solution is to improve the ant colony state transition rules on the basis of the traditional ant colony algorithm, so that the ant colony can select the next point When considering the urgency of station material demand and carbon emissions, it is more in line with the actual production situation. (3) The patent document CN113159687A proposes a material distribution path planning method and system for workshop AGV-UAV collaboration. Its main technical solution is to use an improved genetic algorithm to solve the optimization model of path planning to improve distribution efficiency and so on.

Prior art has following shortcoming:

(1) The assembly line material distribution method described in the patent document CN113344487A requires the feeder to complete the material sorting and assembly in the online edge integrated supermarket according to the material distribution scheduling plan in advance, and store them in the material box, but when it is necessary to distribute materials for the station , the feeder takes out several material boxes and places them in each station in sequence. The disadvantage is that the timeliness and accuracy of material distribution need to be guaranteed by the delivery workers. (2) The patent document CN113947310A builds a relevant mathematical model with the least number of vehicles used and the shortest vehicle driving distance as the optimization objectives. However, in actual production, most production workshops will focus on the shortest completion time to ensure production quality, so the mathematical model is somewhat divorced from actual production. (3) The patent document CN113159687A takes the energy consumption of the vehicle distribution process as the optimization target, but the actual production does not take this as the single optimization target, and the actual reference value is small. At the same time, for areas with complex terrain in the workshop, it is proposed to manually use drones to cooperate with material distribution, which increases the complexity.

Contents of the invention

Aiming at the deficiencies in the above-mentioned prior art, the present invention provides a multi-station, multi-material task assignment and path planning method based on genetic algorithm, aiming at the distribution of multi-station requirements and multi-material in discrete assembly workshops, which takes a long time and is cumbersome and restrictive. Issues such as the completion of assembly tasks allow multiple stations to obtain the correct type of materials within the correct time window, and on the basis of completing material distribution, make the material distribution time as short as possible and improve workshop efficiency.

In order to achieve the above object, the present invention provides a kind of multi-station multi-material task distribution and path planning method based on genetic algorithm, comprising steps:

S1: preprocessing step of material distribution;

The S1 step further comprises the steps of:

S11: Modular design steps;

S12: material clustering step;

S13: key material screening step;

S14: Taking into account the needs of the workstations, the batch delivery step is performed at the wrong time;

S2: Establish a mathematical model with the shortest completion time as the optimization goal;

S3: Solve the mathematical model by using a genetic algorithm according to the constraints.

Preferably, in the step S11: according to the different requirements and common needs of different customers, the work of the entire mechanical equipment is divided into four modules when designing the work: basic module, optional module, customized module and auxiliary module.

Preferably, in the step S12: according to the fixed assembly process of mechanical equipment, using the characteristics of different time windows for each material requirement of each station and fixed material requirements at each stage as the main classification index, the materials are classified according to The time period is divided into four categories, which include: basic, optional, customized and auxiliary.

Preferably, in the step S13: setting up evaluation indexes with different weights to screen key materials, taking into account the economy of the materials, the volume of the materials and the modules in which the materials are located.

Preferably, in the step S14: use the time window to describe the order and urgency of the materials required by the station, assign a distribution number in combination with the module where the materials are located, and assign an initial value to each transport vehicle, And look for the next more urgent station or the more urgent material for delivery, to achieve the shortest optimal path with multiple indicators, and complete the task jointly by multiple delivery sub-paths; the indicators include distance, waiting time and penalty time.

Preferably, in the step S2: the mathematical model is partially simplified according to the actual situation of the delivered materials.

Preferably, the S3 step further comprises the steps of:

S31: Encoding and decoding the mathematical model by using a multi-layer encoding method based on natural numbers;

S32: Take the reciprocal of the total completion time of the delivery plan represented by each chromosome as the fitness value of the genetic algorithm;

S33: Randomly generate an initial population, and use a selection method combining the roulette method and the elite retention strategy to select the initial population;

S34: performing a crossover operation by using a consistent crossover operator and performing a mutation operation by randomly exchanging gene bits;

S35: Eliminate individuals with low fitness, and obtain the optimal solution after multiple rounds of evolution.

The present invention has the following beneficial effects due to the adoption of the above technical scheme:

The present invention first clusters all materials into four modules according to the modular working mode adopted in the design work of mechanical equipment and uses the K-means algorithm to improve the traditional ABC material classification criteria; then uses three screening indicators to select key materials Focus on attention; finally, the genetic algorithm is used to solve the established mathematical model with the shortest delivery time as the main optimization goal, and a dual delivery route with station order and material type is obtained. The problem of multi-station and multi-material distribution in discrete assembly workshops is solved. Compared with the existing technology, the material distribution solution is added, and the practicability of actual production is also taken into account.

Description of drawings

FIG. 1 is a flow chart of a genetic algorithm-based multi-station and multi-material task assignment and path planning method according to an embodiment of the present invention.

Detailed ways

Below according to accompanying drawing Fig. 1, provide the preferred embodiment of the present invention, and describe in detail, enable the function, characteristic of the present invention to be better understood.

Please refer to Fig. 1, a kind of multi-station and multi-material task distribution and path planning method based on genetic algorithm of the embodiment of the present invention, first all materials are improved by using K-means algorithm according to the modular working mode adopted in the design work of mechanical equipment The traditional ABC material classification criteria are used to cluster into four modules; then three screening indicators are used to select the key materials to focus on; finally, the genetic algorithm is used to solve the established mathematical model with the shortest delivery time as the main optimization goal, and the result is Output the distribution route with station sequence and material type; including steps:

S1: preprocessing step of material distribution;

The assembly work of large and medium-sized fluid mechanical equipment is the most important, and how to make materials arrive accurately and in time becomes the key. Therefore, the three tasks before material distribution consist of pre-processing, namely modular design, material clustering and key material screening, and batch-time delivery in consideration of station requirements.

The S1 step further comprises the steps of:

S11: Modular design steps;

According to the different requirements and common needs of different customers, the work of the entire mechanical equipment is divided into four modules during the design work: basic module, optional module, customized module and auxiliary module.

S12: material clustering step;

According to the fixed assembly process of mechanical equipment, using the characteristics of different time windows for each material requirement of each station and the fixed material requirement of each stage as the main classification index, the materials are divided into four categories according to the different time periods. Includes: Basic, Optional, Custom and Auxiliary.

S13: key material screening step;

Not all materials need to be delivered by a special person. At the same time, the logistics resources of the workshop are limited and cannot cover all materials. Therefore, evaluation indicators with different weights are set up to screen key materials, and the economy of materials, material volume and the modules in which materials are located are considered comprehensively.

S14: Taking into account the needs of the workstations, the batch delivery step is performed at the wrong time;

Considering that the modules of the materials required by the stations are roughly the same, but there are differences in types, the time window is used to describe the order and urgency of the materials required by the stations, and the distribution number is allocated according to the modules where the materials are located, and each transport vehicle is assigned an initial Value, and find the next more urgent station or more urgent material for delivery, to achieve the optimal path with the shortest multiple indicators, and multiple distribution sub-paths to complete the task together; indicators include distance, waiting time and penalty time.

S2: Establish a mathematical model with the shortest completion time as the optimization goal;

The material distribution problem in the workshop refers to the planning of the route for the transport material trolley to obtain materials from the distribution center and from the distribution center to each demand station, so that the vehicle can deliver different materials to different stations according to the required path and return to the distribution center. It is to achieve the delivery target while satisfying the demand of the station and satisfying certain conditions and constraints.

The specific implementation process of establishing a mathematical model with the shortest completion time as the optimization goal is as follows:

Assume that the following conditions are met in the multi-station and multi-material distribution of discrete workshops:

1. When all transport vehicles depart from the material supermarket, the actual carrying capacity cannot exceed their maximum carrying capacity;

2. The distance from the material supermarket to any station and the distance between stations are known;

3. During task delivery, the vehicle departs from the material supermarket to each station for material delivery, and when the vehicle returns to the material supermarket, it is considered that a material delivery route is over;

4. A transport vehicle can deliver one or more times to one station or multiple stations;

5. The wake-up and waiting time of the transport vehicle are negligible, the driving speed is constant, and the collision and other situations are not considered;

parameter:

C is the maximum completion time; i/j is the station number being served; k is the number of the transport vehicle; t _i is the time when the transport vehicle arrives at the station; w _i is the necessary waiting time for the service station _i ; The service time of station i; d _ij is the distance between station i and station j; v is the average speed of the car;

Model:

The objective function is as follows:

in:

∑d _ij ≤ L (8);

At the same time, according to the actual situation of the delivered materials, the model is partially simplified as follows:

当取值为1时，表示车辆k负责物料配送从工位i到工位j；

When the value is 1, it means that vehicle k is responsible for material distribution from station i to station j;

当取值为1时，表示车辆k负责配送到达工位j等待装卸物料；

When the value is 1, it means that vehicle k is responsible for delivery to station j and waits for loading and unloading materials;

当取值为1时，表示车辆k负责配送到达工位i进行服务中；

When the value is 1, it means that vehicle k is responsible for delivery to station i for service;

Equation (1) is the objective function with the shortest time to complete the delivery task. It mainly includes vehicle driving time, waiting time at the material supermarket, service time at the station and penalty time for exceeding the set time window. Among them, α and β in formula (2) are the penalty coefficients arriving before and after the time window respectively. Equation (3) ensures that the number of arriving vehicles and leaving vehicles at each station is consistent. Equation (4) is the limitation of the station time window. Equation (5) means that the distribution of all materials needs to be completed within a certain period of time. Equation (6) is the constraint before and after the service time of the vehicle during the process from station i to station j. Equation (7) ensures that the loading capacity of the vehicle is within its maximum load. Equation (8) ensures that the running state of the vehicle does not exceed the furthest running distance of the vehicle.

S3: According to the constraints, the genetic algorithm is used to solve the mathematical model.

The S3 step further includes the steps of:

S31: Encoding and decoding the mathematical model by adopting a multi-layer encoding method based on natural numbers;

S32: Take the reciprocal of the total completion time of the delivery plan represented by each chromosome as the fitness value of the genetic algorithm;

S33: Randomly generate the initial population, and use the selection method combining the roulette method and the elite retention strategy to select the initial population;

S34: performing a crossover operation by using a consistent crossover operator and performing a mutation operation by randomly exchanging gene bits;

S35: Eliminate individuals with low fitness, and obtain the optimal solution after multiple rounds of evolution.

The present invention has been described in detail above with reference to the embodiments of the accompanying drawings, and those skilled in the art can make various changes to the present invention according to the above description. Therefore, some details in the embodiments should not be construed as limiting the present invention, and the present invention will take the scope defined by the appended claims as the protection scope of the present invention.

Claims (7)

1. A multi-station multi-material task distribution and path planning method based on genetic algorithm, comprising steps: S1: preprocessing step of material distribution; The S1 step further comprises the steps of: S11: Modular design steps; S12: material clustering step; S13: key material screening step; S14: Taking into account the needs of the workstations, the batch delivery step is performed at the wrong time; S2: Establish a mathematical model with the shortest completion time as the optimization goal; S3: Solve the mathematical model by using a genetic algorithm according to the constraints. 2. The multi-station and multi-material task assignment and path planning method based on genetic algorithm according to claim 1, characterized in that, in the S11 step: according to the different requirements and common requirements of different customers, the whole The work of mechanical equipment is divided into four modules: basic module, optional module, customized module and auxiliary module. 3. The multi-station and multi-material task assignment and path planning method based on genetic algorithm according to claim 2, characterized in that, in the step S12: according to the fixed assembly process of mechanical equipment, each The main classification indicators are the different time windows of material requirements and the fixed material requirements of each stage. The materials are divided into four categories according to the different time periods. The categories include: basic, optional, customized and auxiliary. 4. The multi-station and multi-material task assignment and path planning method based on genetic algorithm according to claim 3, characterized in that, in the S13 step: set up evaluation indexes with different weights to screen key materials, and use the materials Consider comprehensively the economy, material volume and the module in which the material is located. 5. The multi-station and multi-material task assignment and path planning method based on genetic algorithm according to claim 4, characterized in that, in the step S14: using the time window to describe the required material of the station The sequence and urgency, combined with the module where the material is located, assigns a distribution number, assigns an initial value to each transport vehicle, and searches for the next more urgent station or more urgent material for distribution, reaching as many as The optimal path with the shortest index is the shortest path, and multiple distribution sub-paths jointly complete the task; the index includes distance, waiting time and penalty time. 6. The multi-station and multi-material task assignment and path planning method based on genetic algorithm according to claim 5, characterized in that, in the S2 step: according to the actual situation of the distribution materials, the mathematical model is partially simplified . 7. The multi-station and multi-material task assignment and path planning method based on genetic algorithm according to claim 6, characterized in that, the S3 step further comprises the steps: S31: Encoding and decoding the mathematical model by using a multi-layer encoding method based on natural numbers; S32: Take the reciprocal of the total completion time of the delivery plan represented by each chromosome as the fitness value of the genetic algorithm; S33: Randomly generate an initial population, and use a selection method combining the roulette method and the elite retention strategy to select the initial population; S34: performing a crossover operation by using a consistent crossover operator and performing a mutation operation by randomly exchanging gene bits; S35: Eliminate individuals with low fitness, and obtain the optimal solution after multiple rounds of evolution.

Images