CN115063064A - A storage space allocation method for production logistics warehouse based on genetic algorithm - Google Patents

Abstract

The present invention provides a method for allocating storage space in a production logistics warehouse based on a genetic algorithm, comprising the following steps: 1) establishing a mathematical model of the storage space in the warehouse: the mathematical model includes a three-dimensional coordinate system and several shelves located in the three-dimensional coordinate system 2) The design of the chromosomal coding of the cargo: a chromosomal coding is composed of multiple genes, and each of the four genes is divided into a group. The first three genes in each group represent the x, y, and z coordinates of a cargo. A gene represents the group to which the goods belong; 3) Fitness function design: design a fitness function according to the storage conditions of the goods in the warehouse; 4) Solve the optimal solution by genetic algorithm, and obtain the optimal warehouse storage space allocation method. The method can solve the technical problems of low warehouse picking efficiency, large moving distance of goods and high warehousing operation cost of the existing warehouse storage space allocation scheme.

Description

A storage space allocation method for production logistics warehouse based on genetic algorithm

technical field

The invention relates to the technical field of warehouse storage space allocation methods, in particular to a production logistics warehouse storage space allocation method based on a genetic algorithm.

Background technique

In the field of production logistics optimization, warehousing location allocation refers to the process that an enterprise performs some inventory optimization or location allocation for its own warehouse according to its own product characteristics, warehouse characteristics, demand and its changing factors. The allocation plan of the cargo space needs to be comprehensively considered and customized according to factors such as warehouse characteristics, product characteristics, and cargo space planning, so as to achieve the goals of improving warehouse picking efficiency, reducing the moving distance of goods, and reducing warehouse operation costs.

SUMMARY OF THE INVENTION

The technical problem to be solved by the present invention is to provide a production logistics warehouse storage space allocation method based on genetic algorithm, so as to solve the technical problems of low warehouse picking efficiency, large moving distance of goods and high storage operation cost of the existing warehouse storage space allocation scheme .

In order to solve the above-mentioned technical problems, the technical scheme of the present invention is: a method for allocating storage positions in a production logistics warehouse based on a genetic algorithm, comprising the following steps:

1) establish a warehouse storage location mathematical model: the mathematical model includes a three-dimensional coordinate system and several shelves located in the three-dimensional coordinate system;

2) Cargo chromosome coding design: a chromosome code is composed of multiple genes, each of the four genes is divided into a group, the first three genes in each group represent the x, y, z coordinates of a cargo, and the last one Gene indicates the group to which the goods belong;

3) Fitness function design: according to the storage conditions of the goods in the warehouse, the fitness function is designed;

4) Use the fitness function of step 3) to solve the optimal solution through the genetic algorithm, and obtain the optimal warehouse storage space allocation method.

As an improved way, in step 3), the fitness function is required to be set in combination with the following conditions:

1) The goods with high frequency are placed close to the origin;

2) The goods that are often in and out of the warehouse are stored together;

3) Put heavy goods on the bottom of the shelf to ensure the vertical stability of the shelf.

As an improved way, in step 2), the weight coefficient is used to convert the multi-objective model into a single-objective model to design a fitness function.

As an improved way, in the process of selecting each individual in the group, the fitness value of similar individuals is reduced by the niche technology, which specifically includes the following steps:

1) Set a niche radius σ _share ;

2) Check the Hamming or Euclidean distance d _ij of pairs of individuals of the population;

3) When d _ij <σ _share , penalize the fitness of one of the individuals to distinguish the two individuals, so that the individual with small fitness will be eliminated by the subsequent selection operator with a high probability.

The technical effects obtained by adopting the above technical solutions are:

The warehouse storage space allocation method of the present application is optimized according to the frequency of in and out of the warehouse, the correlation and stability of the goods, and the goods that are in and out of the warehouse with high frequency can be preferentially placed near the entrance and exit of the warehouse, and the goods that are often in and out of the warehouse together are stored together, so as to reduce the cost of storage. The in-out distance of high-frequency inbound and outbound goods reduces the effect of warehouse operation time. The distribution method of the present application can also make the layout of the goods on the shelf more reasonable, and improve the stability of the shelf after the goods are stored.

Since in the process of selecting each individual in the group, the fitness value of similar individuals is reduced by the niche technology, and the individuals in the population will be effectively separated through the genetic algorithm of the niche process, that is, within a σ There will only be a small number of individuals in the niche within the _share radius, so it has a better effect on maintaining the diversity of the population, which can speed up the convergence speed of the genetic algorithm.

Description of drawings

The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

Fig. 1 is the mathematical model diagram of warehouse storage position;

Figure 2 is a flow chart of the basic genetic algorithm;

Detailed ways

A method for allocating storage space in a production logistics warehouse based on a genetic algorithm, comprising the following steps:

1. Establish a mathematical model of warehouse storage location: the mathematical model includes a three-dimensional coordinate system and several shelves located in the three-dimensional coordinate system, as shown in FIG. 1 .

2. Cargo chromosome coding design: a chromosome code is composed of multiple genes, and each of the four genes is divided into a group. The first three genes in each group represent the x, y, and z coordinates of a cargo, and the last one Gene indicates the group to which the item belongs.

3. Fitness function design: According to the mathematical model of the warehouse established in step 1, use x _i , y _i , z _i to represent the coordinates on the three axes of x, y, and z of a product. And use qi to represent the average coordinates of the i-th group of goods, use d to represent the sum of the average coordinates of all goods to its own group, h to represent the height of each cargo space, and l to represent the distance between shelves. Use _Pi to represent the turnover rate of each shipment, that is, the frequency of use. M _i represents the weight of each cargo. N represents the quantity of all items, and n represents the quantity of item groupings. Use W to denote the product of the center and distance of all item weights in the z-direction.

The fitness function is required to be set in combination with the following conditions:

(1) The goods with high frequency of use are placed as close to the origin as possible. To optimize according to this goal, the priority is to place the goods with high frequency in and out of the warehouse near the entrance and exit of the warehouse, so as to reduce the in-out distance of the goods in and out of the warehouse and reduce the warehouse. Goals such as work hours.

To this end, the design objective function is:

According to the above formula, it can be seen that the farther the coordinates of the goods are from the origin, the larger the F ₁ ; on the contrary, the closer the coordinates of the goods are to the origin of the coordinates, the smaller the objective function F ₁ is. Similarly, the higher the turnover rate is. Tall items should be placed at the origin as much as possible to reduce the value _of the objective function F1. Therefore, in the process of optimizing this objective function, the effect of placing the goods with high turnover rate at the origin as much as possible will be achieved, thereby speeding up the high turnover rate, the efficiency of goods in and out of the warehouse, and shortening the operation time.

In order to let the genetic algorithm optimize multiple objectives at the same time without being affected by the different dimensions of the objective function, combined with the convention of optimizing the objective by maximizing the genetic algorithm, we made the objective function established above into the following formula The conversion shown:

(2) The goods that are often put in and out of the warehouse together should be stored together as much as possible, that is, the optimization should be carried out according to the correlation of the goods, which means that in the warehouse operation of the enterprise, if there are some goods that are often put in and out of the warehouse together, then the goods with high correlation Putting them together can achieve the goals of facilitating staff in and out of the warehouse and shortening the warehouse operation time and operation distance. Optimizing according to the correlation of the goods is to establish an appropriate objective function, and try to achieve the goal of placing the goods that are often in and out of the warehouse together in a nearby location.

First find the average coordinates within the group for each group of items:

Among them, i represents the i-th group; j represents the j-th item in a group; _ki represents the quantity of the i-th group of goods.

Then, calculate the sum of the distances of all items to the average coordinates within their respective groups:

Finally, in order to avoid the offset of the optimization objective caused by the difference of dimensions between multiple objectives, and combined with the habit of maximizing the objective function of the genetic algorithm, we make a change to this distance to obtain the final objective function:

(3) Put heavy goods on the bottom of the shelf to ensure the vertical stability of the shelf, which means that the enterprise considers the structure of its own warehouse shelves and the weight and shape of the goods, so that the goods can be stored in the warehouse. Guarantee the stability of the shelf. In terms of shelf stability, there are often two optimization directions. One is the shelf stability in the horizontal direction, that is, to ensure that the weight of the goods on the left and right sides of the shelf is not much different in the process of cargo space allocation. The second is the shelf stability in the vertical direction, that is, to ensure that heavy objects are placed on the bottom of the shelf during the allocation process.

Design objective function:

Among them, _zi represents the z-axis coordinate of the ith cargo.

In order to ensure that the optimization process of the genetic algorithm is not affected by the dimensions of different objective functions, and at the same time satisfy the maximization objective function of the genetic algorithm, the formula is rewritten:

The following weight coefficients are used to transform the multi-objective model into a single-objective model to design the fitness function.

maxf=w ₁ ×F ₁ +w ₂ ×F ₂ +w ₃ ×F ₃

Among them: w _i represents the weight value of the i-th objective, and F _i is the corresponding i-th objective function. The weight value of the objective function is evaluated according to the factory's experience in outgoing goods.

4. Use the fitness function of step 3 to solve the optimal solution through the genetic algorithm (the steps of the genetic algorithm are shown in Figure 2), and obtain the optimal warehouse storage space allocation method.

As a preferred embodiment, in the process of selecting each individual in the group, the fitness value of similar individuals is reduced by the niche technology, including the following steps:

1) Set a niche radius σ _share ;

2) Check the Hamming or Euclidean distance d _ij of pairs of individuals of the population;

3) When d _ij <σ _share , penalize the fitness of one of the individuals to distinguish the two individuals, so that the individual with small fitness will be eliminated by the subsequent selection operator with a high probability.

After the genetic algorithm of the niche process, the individuals in the population will be effectively separated, that is, there will only be a small number of individuals in a niche within a σ _share radius, so it has a better effect on maintaining the diversity of the population, so that it can be Speed up the convergence of genetic algorithms.

The warehouse storage space allocation method of the present application is optimized according to the frequency of in and out of the warehouse, the correlation and stability of the goods, and the goods that are in and out of the warehouse with high frequency can be preferentially placed near the entrance and exit of the warehouse, and the goods that are often in and out of the warehouse together are stored together, so as to reduce the cost of storage. The in-out distance of high-frequency inbound and outbound goods reduces the effect of warehouse operation time. The distribution method of the present application can also make the layout of the goods on the shelf more reasonable, and improve the stability of the shelf after the goods are stored.

Claims (4)

1. a production logistics warehouse storage location allocation method based on genetic algorithm, is characterized in that, comprises the steps: 1) establish a warehouse storage location mathematical model: the mathematical model includes a three-dimensional coordinate system and several shelves located in the three-dimensional coordinate system; 2) Cargo chromosome coding design: a chromosome code is composed of multiple genes, each of the four genes is divided into a group, the first three genes in each group represent the x, y, z coordinates of a cargo, and the last one Gene indicates the group to which the goods belong; 3) Fitness function design: according to the storage conditions of the goods in the warehouse, the fitness function is designed; 4) The optimal solution is solved by genetic algorithm, and the optimal warehouse storage space allocation method is obtained. 2. the production logistics warehouse storage location allocation method based on genetic algorithm as claimed in claim 1, is characterized in that, in step 3), in conjunction with following condition requirements, set fitness function: 1) The goods with high frequency are placed close to the origin; 2) The goods that are often in and out of the warehouse are stored together; 3) Put heavy goods on the bottom of the shelf to ensure the vertical stability of the shelf. 3. The production logistics warehouse storage location allocation method based on genetic algorithm as claimed in claim 2, is characterized in that, in step 4), utilize weight coefficient to convert multi-objective model into single-objective model to design fitness function. 4. The method for allocating storage space in a production logistics warehouse based on a genetic algorithm as claimed in claim 1, wherein in the process of selecting each individual in the group, the fitness value of the similar individual is reduced by the niche technology , which includes the following steps: 1) Set a niche radius σ _share ; 2) Check the Hamming or Euclidean distance d _ij of pairs of individuals of the population; 3) When d _ij <σ _share , penalize the fitness of one of the individuals to distinguish the two individuals, so that the individual with small fitness will be eliminated by the subsequent selection operator with a high probability.

Images