A workshop intelligent distributor model and its solution algorithm based on Internet of Things technology

Abstract

The present invention discloses a workshop intelligent distributor model based on the Internet of Things technology and its solution algorithm. Through the analysis of the logistics distribution mode of the manufacturing workshop, a multi-index workshop intelligent distributor mathematical model is established with the decision-making objectives of maximizing material distribution efficiency, maximizing the packing load rate, minimizing the distribution distance, and minimizing the time loss in distribution, providing a systematic and optimized distribution strategy for material distribution, and then using a maximum flush loading heuristic algorithm to solve the packing problem, and designing a genetic algorithm as the main solution algorithm of the model, and verifying the scientificity and effectiveness of the algorithm through examples. On the basis of the Internet of Things technology, the present invention uses RFID technology to realize intelligent perception and monitoring of material information, builds a workshop WLAN network, and develops a smart phone App based on Android 4.0 as the carrier of the algorithm, and applies the Internet of Things technology to workshop logistics distribution, so as to realize on-time, on-site, on-demand and efficient distribution of workshop materials.

Description

Technical Field

[0001] The present invention belongs to the technical field of Internet of Things perception and material distribution, and specifically relates to a workshop intelligent distributor model and its solution algorithm based on Internet of Things technology. Background Art

[0002] As competition in the modern manufacturing market becomes increasingly fierce, most domestic manufacturing industries adopt a flexible production model with multiple varieties and small to medium batches. However, domestic small and medium-sized enterprises have yet to achieve intelligent and digital transformation. Consequently, traditional manufacturing workshops lack effective methods for obtaining real-time material demand information, are unable to effectively plan distribution routes, and are unable to optimize the assignment of logistics personnel or vehicles. This often leads to material delivery delays and errors, resulting in production waits or even interruptions, increased backlogs in lineside warehouses, and reduced production efficiency. To unify workshop material delivery with production operations, improve the accuracy and punctuality of material delivery, effectively reduce lineside inventory, and achieve visual management of logistics information, this paper proposes a workshop intelligent distributor model and its solution algorithm based on Internet of Things technology. By constructing an intelligent distributor model suitable for workshop logistics, a genetic algorithm is embedded in the heuristic algorithm for solving the three-dimensional packing problem to solve the workshop intelligent distributor model and provide an optimal distribution solution. Summary of the Invention

[0003] To address the characteristics of flexible production models with high-variety, small- to medium-sized batches, mathematical modeling of a workshop intelligent distributor was developed from four perspectives: assignment, packing, vehicle routing optimization, and delivery departure time determination. The model was then integrated using a multi-metric comprehensive evaluation approach. A genetic algorithm was then used as the main algorithm to solve the workshop intelligent distributor model, while a heuristic algorithm based on the maximum flush loading algorithm was employed to solve the packing problem within the model. The algorithm's scientificity and superiority were verified using a numerical example. Finally, the workshop intelligent distributor was constructed using Internet of Things (IoT) technology. RFID technology enabled real-time identification and sensing of material information. A workshop WLAN network was established to transmit and share logistics data. An Android 4.0 smartphone app was developed as the algorithm's platform, enabling precise control of the material distribution process.

[0004] To achieve the above object, the present invention adopts the following technical solutions:

[0005] The present invention provides a workshop intelligent dispenser model based on Internet of Things technology and its solution algorithm, comprising the following steps:

[0006] (1) Readers are installed at the exit of the turnover warehouse, at each workstation of the assembly line, and at the entrance of the finished product warehouse. Electronic tags are installed on the turnover boxes containing small materials, large materials, and finished products, respectively, to achieve unique identification of materials in the workshop. The information of each material is contained in its electronic tag. RFID technology is used to quickly and accurately identify materials, establish a workshop WLAN network, and realize visual management through smartphone apps.

[0007] (2) Establish a workshop intelligent distributor model and calculate the optimal distribution plan based on the currently idle material distributors and logistics transportation equipment and the list of tasks that need to be distributed through algorithms.

[0008] (3) The assigned material delivery personnel receive reminders from the mobile phone APP and go to the turnover warehouse to pick the corresponding materials according to the material list given in the APP and pack them according to the packing plan; the materials are shipped out through the RFID equipment of the turnover warehouse; when the departure time of the material delivery personnel arrives, the App will sound a reminder and the material delivery personnel will start delivery; the material delivery personnel will check the logistics delivery route on the mobile phone App and deliver the materials in the order of delivery tasks according to the prompts of the App; when the materials arrive at the workstation, the RFID equipment in the workstation senses the quantity of materials and transmits it to the terminal database in real time; the material delivery personnel will return to the turnover warehouse after completing the delivery tasks in the current delivery task list; after the production line assembly is completed, the finished products are transported to the finished product warehouse and pass through the RFID equipment at the entrance of the finished product warehouse.

[0009] Furthermore, in step (1), the workstation data of the production site is read according to the IoT related equipment, mainly including the workstation number, required material number, required material quantity, etc., to generate the workstation demand and add it to the current demand list in the order of the latest delivery time of the materials.

[0010] Furthermore, the intelligent workshop logistics distributor model in step (2) is constructed based on four decision-making objectives: improving distribution efficiency, packing load rate, shortening distribution distance, and minimizing the total time loss in distribution. Therefore, the information of the workshop intelligent distributor model is as follows:

[0011] ①Parameter list:

[0012] Define the meaning of the following symbols:

[0013] L——the length of the turnover box

[0014] W——Width of turnover box

[0015] H——the height of the turnover box

[0016] G——maximum load capacity of turnover box

[0017] N——Number of material types

[0018] K——number of workstations

[0019] W——Number of forklifts

[0020] G W ——Maximum load of each vehicle

[0021] P——Production line production rhythm, which should be measured according to the actual situation of the production line

[0022] T min ——The minimum time that the current line inventory margin can ensure the normal operation of the production line

[0023] n——Logistics deliveryman

[0024] m——number of logistics equipment

[0025] TS——Current number of delivery tasks

[0026] v i ——The average delivery speed of material delivery personnel (1≤i≤n, and i∈Z)

[0027] v j ——The operating speed of logistics equipment (1≤j≤m, and j∈Z)

[0028] v ij ——The ability of material delivery person i to work with logistics equipment j

[0029] b i ——The maximum number of delivery tasks that deliveryman i can complete

[0030] S ts ——The delivery path distance selected by the system for the current delivery task ts

[0031] l——Material

[0032] k——workstation

[0033] w——Forklift

[0034] u——delivery order

[0035] l l ——The length of the lth type of material (1≤l≤N, and l∈Z)

[0036] w l ——Width of the lth type of material (1≤l≤N, and l∈Z)

[0037] h l ——Height of the lth type of material (1≤l≤N, and l∈Z)

[0038] g l ——Weight of material of category l (1≤l≤N, and l∈Z)

[0039] q l ——Number of materials in category l (1≤l≤N, and l∈Z)

[0040] num l ——The quantity of the first type of materials loaded into the turnover box (1≤l≤N and l∈Z,0≤num l ≤q l and num l ∈Z)

[0041] BWP l ——Height of the center of gravity of the first type of material (1≤l≤N, and l∈Z)

[0042] d kk′ ——The distance between station k and station k′

[0043] m k ——Total weight of materials required for the kth station

[0044] BT kl ——The earliest time in the hard time window for the delivery of material l required by workstation k

[0045] LT kl ——The latest time that material l arrives at workstation k

[0046] RT kl ——The time when workstation k sends the material l demand instruction

[0047] IL kl ——The current inventory quantity of line material l at workstation k

[0048] TL kl ——Specific quantity of material l delivered to workstation k

[0049] r kl ——Consumption rate of material l in workstation k

[0050] SS kl ——Safety stock of material l at workstation k

[0051] t kw ——The time it takes for forklift w to reach delivery station k

[0052] [k]——the kth workstation in the delivery task list, k=0,1,2,…K;

[0053] ——The time it takes for material l to arrive at the required workstation during the uth delivery by the material distributor

[0054] ——The time when the material delivery person leaves the workstation where material l is located during the uth delivery. When l = 0, this variable represents the time when the material delivery person starts loading the material in the turnover warehouse.

[0055] T S (l)——The loading time of material l in the turnover bin

[0056] T u (l)——the unloading time of material l at the required station

[0057] T kk′ (u)——The time it takes for the material delivery person to go from workstation k to workstation k′ in the uth delivery. When k=0 and k′=0, it means that the material delivery person is in the turnover warehouse.

[0058] P k (t): Time loss function value of material delivery to workstation k at time t

[0059] Decision variables

[0060] Decision variables

[0061] Decision variables

[0062] constant λ (0≤λ≤1)

[0063] constants k1, k2, k3 (k1+k2+k3=1 and k1, k2, k3∈[0,1])

[0064] ②Model assumptions

[0065] 1. Unify logistics containers and use turnover boxes of the same size during distribution;

[0066] 2. Simplify materials of different shapes and specifications into the smallest rectangular shape that meets their size;

[0067] 3. The location of the turnover warehouse is determined and there are sufficient and reasonable number of forklifts to distribute materials;

[0068] 4. The turnover warehouse has sufficient materials and can meet the production needs of all workstations;

[0069] 5. Assuming that all forklifts have the same speed, the maximum distance a forklift can travel is infinite;

[0070] 6. Since the forklifts have the same speed, under the premise of meeting the material needs of the workstation, the forklift travels the shortest distance and the time spent on delivery must be the shortest, that is, the total cost is the lowest.

[0071] ③Model integration

[0072] The integrated model should take into account that during the packing process, the order in which materials are packed should be the opposite of the order in which they are unloaded.

[0073]

[0074] ST

[0075]

[0076]

[0077]

[0078]

[0079]

[0080]

[0081]

[0082]

[0083] v ij =λv i +(1-λ)v j (10)

[0084]

[0085]

[0086]

[0087] F=k1CR+k2WR+k3CG (14)

[0088]

[0089]

[0090]

[0091] k1+k2+k3=1 and k1, k2, k3∈[0,1] (18)

[0092]

[0093]

[0094]

[0095]

[0096]

[0097] BT kl ≤t kw ≤LT kl (k=1,2,…,K) (24)

[0098]

[0099]

[0100]

[0101]

[0102]

[0103] Wherein, expression (1) represents the comprehensive evaluation index of the multi-index comprehensive evaluation method;

[0104] Expression (2) represents the weight of the multi-index comprehensive evaluation method;

[0105] Expression (3) represents the standard deviation of the dimensionless standard value of each indicator;

[0106] Constraint (4) limits the range of weight values;

[0107] Expressions (5) to (8) respectively represent the time required to complete the current delivery task, the space utilization rate, weight loading rate and overall center of gravity position of the integrated packaging, the length of the delivery path and the total time loss of one delivery.

[0108] Expression (9) indicates the calculation method of the time consumed to complete the current delivery task;

[0109] Expression (10) indicates the calculation method of the current working capacity of logistics personnel and logistics equipment;

[0110] Constraints (11) to (13) indicate that any delivery task is completed by and only by one material delivery person, and the number of delivery tasks that some material delivery persons may undertake does not exceed the upper limit of their own work capacity;

[0111] Expression (14) indicates the calculation method of the comprehensive packing space utilization, weight loading rate and overall center of gravity position index;

[0112] Expressions (15) to (17) respectively give the calculation methods of the functions expressing the packing space utilization, weight loading rate and overall center of gravity position;

[0113] Constraint (18) limits the range of values ​​of the constant weight;

[0114] Expression (19) indicates the calculation method of the delivery path length;

[0115] Constraint (20) indicates that the total material demand of all workstations on each delivery route does not exceed the carrying capacity of the forklift;

[0116] Constraint (21) means that the materials at each workstation are delivered by one and only one forklift;

[0117] Constraints (22) and (23) indicate that each forklift must depart from the turnover warehouse and return to the turnover warehouse after completing the delivery task;

[0118] Constraint (24) indicates that the time it takes for the forklift to deliver the material to the workstation meets the time window requirement;

[0119] Expression (25) indicates the calculation method of the time it takes for the forklift to arrive at the delivery station;

[0120] Expressions (26) to (28) indicate the earliest time of the hard time window required by workstation k to deliver material l, the latest time for material l to arrive at workstation k, and the calculation method of the safety stock of material l at workstation k.

[0121] Expression (29) indicates the calculation method of the total time loss of a delivery.

[0122] Furthermore, the genetic algorithm designed in step (2) specifically includes the following steps:

[0123] The first step is to establish four priority rules, logistics and transportation equipment status and four information tables

[0124] (1) First, other constraints are considered only if the current remaining capacity of the logistics equipment can fully load all the required materials at the required delivery stations;

[0125] (2) First, satisfy the time window constraints of the model to ensure that the materials are delivered to the required workstations within the appropriate time period;

[0126] (3) Secondly, the shortest path constraint of the model is satisfied. Under the premise of meeting the time constraint, the logistics equipment delivers the materials to the required workstation along the shortest path;

[0127] (4) Finally, if a workstation cannot meet its distribution needs in the current batch because its suitable distribution time period is significantly different from that of other workstations, or the current logistics and transportation equipment is not suitable for loading the required materials, its distribution needs will be automatically included in the next distribution batch.

[0128] The status of logistics and transportation equipment (such as forklifts, etc.) is used to advance the operation of the algorithm. The status of logistics and transportation equipment is:

[0129] (1) Assignment status (whether assigned);

[0130] (2) Loading status (whether loading is complete (not necessarily full));

[0131] (3) departure status (whether it has departed for delivery);

[0132] (4) Delivery status (whether the workstation requirements of the delivery list are met).

[0133] The algorithm involves four lists: (1) a list of available logistics and transportation equipment and material delivery personnel; (2) a list of workstation requirements; (3) a delivery list; and (4) a list of loadable items in the packing problem (i.e., a list of what types of materials can be loaded into a turnover box). The information required by the workstation in the workstation requirement list includes: (1) the workstation number; (2) the number of the material required by the workstation; and (3) the quantity of the material required by the workstation. This information is provided by IoT-related devices after reading the workstation data at the production site. The workstation requirement list and the current delivery list are in an inclusive relationship. A delivery list is completed by one logistics delivery device and one material delivery person.

[0134] In the second step, the genetic algorithm is used as the main algorithm to solve the workshop intelligent distributor model. The specific process is as follows:

[0135] 1. Encoding and decoding: Based on various encoding methods, a multi-part encoding method is adopted, dividing the gene into four parts. The encoded genes use integers. The first part is the number of the material delivery person, the second part is the logistics transportation equipment number, the third part is the workstation delivery sequence, and the fourth part is the departure time of the material delivery person.

[0136] 2. Fitness function evaluation: take the inverse of the objective function of the workshop intelligent distribution model as its fitness function:

[0137]

[0138] 3. Roulette is used in the selection of basic operations

[0139] The roulette wheel is calculated in the algorithm as follows:

[0140]

[0141]

[0142]

[0143]

[0144] Formula (a) calculates the fitness of a single chromosome, formula (b) calculates the sum of the fitness of all individuals in the contemporary population, formula (c) calculates the selection probability of a single chromosome, and formula (d) calculates the cumulative probability. A random number r∈[0,1] is generated. If r∈[Q k-1 ,Q k ], then select individual x k , k∈(1,2,3…pop_size).

[0145] The basic crossover operation divides the chromosome into three parts. The first part is the delivery driver's number and the logistics equipment number; the second part is the delivery order of the workstation; and the third part is the delivery driver's departure time. Mutation is similarly divided into three parts. Through these operations, new offspring are generated. Finally, through continuous selection and iteration, the optimal complete logistics delivery solution is obtained.

[0146] Beneficial effects of the present invention:

[0147] 1. The present invention combines the current manufacturing industry's characteristics of producing multiple varieties and small batches, analyzes its operation workshop logistics, and solves the current problem of difficult evaluation indicators for workshop logistics benefits by conducting a multi-index comprehensive evaluation of the four indicators of improving distribution efficiency, packing load rate, shortening distribution distance, and minimizing total time loss in distribution. It proposes an intelligent workshop distributor model, which is effectively used to evaluate the multi-index benefits of workshop logistics distribution.

[0148] 2. The present invention uses the genetic algorithm as the main body and embeds the maximum flush loading algorithm to solve the workshop intelligent distributor model. The advantages of the genetic algorithm with good adaptability and convergence are well applied to this problem, which solves the shortcomings of the traditional workshop push-type distribution, reduces logistics costs, saves human resources, and improves the production efficiency of the enterprise. BRIEF DESCRIPTION OF THE DRAWINGS

[0149] Figure 1 This is the technical roadmap of the present invention.

[0150] Figure 2 Flowchart of the genetic algorithm for solving this problem. DETAILED DESCRIPTION

[0151] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to examples and drawings. The contents mentioned in the embodiments are not intended to limit the present invention.

[0152] A workshop intelligent distributor model based on Internet of Things technology. The modeling process is as follows:

[0153] S1: Determine the decision goal

[0154] By using the average delivery speed, packing load rate, vehicle route distance and total time loss during the delivery process as decision-making targets, the total time loss during delivery is reduced to a minimum, providing a systematically optimized delivery strategy for material delivery;

[0155] Relevant parameter information is obtained during workshop logistics and distribution. Readers and writers are installed at the exit of the turnover warehouse, at each workstation on the assembly line, and at the entrance of the finished product warehouse. Electronic tags are installed on turnover boxes containing small materials, large materials, and finished products, respectively. This uniquely identifies materials in the workshop. The information of each material is contained in its electronic tag, and RFID technology is used to quickly and accurately identify materials. The IoT device reads the workstation data at the production site, mainly including the workstation number, the required material number, and the required material quantity. The workstation requirements are then generated and entered into the current requirements list in order of the latest material delivery time.

[0156] In the workshop logistics distribution, the corresponding parameters are obtained through the IoT devices, and the parameter symbols are defined as follows:

[0157] L——the length of the turnover box

[0158] W——Width of turnover box

[0159] H——the height of the turnover box

[0160] G——maximum load capacity of turnover box

[0161] N——Number of material types

[0162] K——number of workstations

[0163] W——Number of forklifts

[0164] G W ——Maximum load of each vehicle

[0165] P——Production line production rhythm, which should be measured according to the actual situation of the production line

[0166] T min ——The minimum time that the current line inventory margin can ensure the normal operation of the production line

[0167] n——Logistics deliveryman

[0168] m——number of logistics equipment

[0169] TS——Current number of delivery tasks

[0170] v i ——The average delivery speed of material delivery personnel (1≤i≤n, and i∈Z)

[0171] v j——The operating speed of logistics equipment (1≤j≤m, and j∈Z)

[0172] v ij ——The ability of material delivery person i to work with logistics equipment j

[0173] b i ——The maximum number of delivery tasks that deliveryman i can complete

[0174] S ts ——The delivery path distance selected by the system for the current delivery task ts

[0175] l——Material

[0176] k——workstation

[0177] w——Forklift

[0178] u——delivery order

[0179] l l ——The length of the lth type of material (1≤l≤N, and l∈Z)

[0180] w l ——Width of the lth type of material (1≤l≤N, and l∈Z)

[0181] h l ——Height of the lth type of material (1≤l≤N, and l∈Z)

[0182] g l ——Weight of material of category l (1≤l≤N, and l∈Z)

[0183] q l ——Number of materials in category l (1≤l≤N, and l∈Z)

[0184] num l ——The quantity of the first type of materials loaded into the turnover box (1≤l≤N and l∈Z,0≤num l ≤q l and num l ∈Z)

[0185] BWP l ——Height of the center of gravity of the first type of material (1≤l≤N, and l∈Z)

[0186] d kk′ ——The distance between station k and station k′

[0187] m k ——Total weight of materials required for the kth station

[0188] BT kl ——The earliest time in the hard time window for the delivery of material l required by workstation k

[0189] LT kl ——The latest time that material l arrives at workstation k

[0190] RT kl ——The time when workstation k sends the material l demand instruction

[0191] IL kl ——The current inventory quantity of line material l at workstation k

[0192] TL kl ——Specific quantity of material l delivered to workstation k

[0193] r kl ——Consumption rate of material l in workstation k

[0194] SS kl ——Safety stock of material l at workstation k

[0195] t kw ——The time it takes for forklift w to reach delivery station k

[0196] [k]——the kth workstation in the delivery task list, k=0,1,2,…K;

[0197] ——The time it takes for material l to arrive at the required workstation during the uth delivery by the material distributor

[0198] ——The time when the material delivery person leaves the workstation where material l is located during the uth delivery. When l = 0, this variable represents the time when the material delivery person starts loading the material in the turnover warehouse.

[0199] T S (l)——The loading time of material l in the turnover bin

[0200] T u (l)——the unloading time of material l at the required station

[0201] T kk′ (u)——The time it takes for the material delivery person to go from workstation k to workstation k′ in the uth delivery. When k=0 and k′=0, it means that the material delivery person is in the turnover warehouse.

[0202] P k (t): Time loss function value of material delivery to workstation k at time t

[0203] Decision variables

[0204] Decision variables

[0205] Decision variables

[0206] constant λ (0≤λ≤1)

[0207] constants k1, k2, k3 (k1+k2+k3=1 and k1, k2, k3∈[0,1]);

[0208] S2: The workshop intelligent dispatcher model must meet the following constraints:

[0209] 1. Unify logistics containers and use turnover boxes of the same size during distribution;

[0210] 2. Simplify and mark materials of different shapes and specifications into the smallest rectangular shape that meets their size;

[0211] 3. The location of the turnover warehouse is determined and equipped with sufficient forklifts to distribute materials;

[0212] 4. The turnover warehouse has sufficient materials and can meet the production needs of all workstations;

[0213] 5. If the speed of all forklifts is fixed and the same, the maximum distance a forklift can travel is infinite;

[0214] 6. Since the forklifts have the same speed, if the forklift travels the shortest distance and meets the material needs of the workstation, the delivery time will be the shortest, that is, the total cost will be the lowest;

[0215] S3: Calculation formula and model integration in the process of establishing the workshop intelligent dispatcher model

[0216] The calculation formula in the process of establishing the workshop intelligent distributor model is as follows:

[0217]

[0218] ST

[0219]

[0220]

[0221]

[0222]

[0223]

[0224]

[0225]

[0226]

[0227] v ij =λv i +(1-λ)v j (10)

[0228]

[0229]

[0230]

[0231] F=k1CR+k2WR+k3CG (14)

[0232]

[0233]

[0234]

[0235] k1+k2+k3=1 and k1, k2, k3∈[0,1] (18)

[0236]

[0237]

[0238]

[0239]

[0240]

[0241] BT kl ≤t kw ≤LT kl (k=1,2,…,K) (24)

[0242]

[0243]

[0244]

[0245]

[0246] In the above calculation formula, formula (1) represents the calculation formula of the comprehensive evaluation index of the multi-index comprehensive evaluation method; formula (2) represents the calculation formula of the weight of the multi-index comprehensive evaluation method;

[0247] Expression (3) represents the calculation formula of the standard deviation of the dimensionless standard value of each indicator;

[0248] Constraint (4) limits the range of weight values ​​in the calculation formula;

[0249] Expressions (5) to (8) respectively represent the calculation formulas of the four indicators: the time consumed to complete the current delivery task, the space utilization rate, weight loading rate and overall center of gravity position of the comprehensive packaging, the length of the delivery path, and the total time loss of one delivery;

[0250] Expression (9) represents the calculation method of the time consumed to complete the current delivery task;

[0251] Expression (10) represents the calculation method of the current working capacity of logistics personnel and logistics equipment;

[0252] Constraint (11) indicates whether to assign a logistics delivery person to complete the task;

[0253] Constraint (12) indicates that any delivery task can be completed by only one material delivery person;

[0254] Constraint (13) means that the delivery tasks undertaken by the material delivery personnel shall not exceed the upper limit of their own work capacity;

[0255] Expression (14) indicates the calculation method of the comprehensive packing space utilization, weight loading rate and overall center of gravity position index;

[0256] Expressions (15) to (17) respectively give the calculation methods of the functions expressing the packing space utilization, weight loading rate and overall center of gravity position;

[0257] Constraint formula (18) limits the value range of the constant weight;

[0258] Expression (19) indicates the calculation method of the delivery path length;

[0259] Constraint formula (20) indicates that the total demand for materials at all workstations on each delivery route does not exceed the carrying capacity of the forklift;

[0260] Constraint formula (21) indicates that the materials at each workstation are delivered by one and only one forklift;

[0261] Constraint (22) means that each forklift needs to be dispatched from the turnover warehouse;

[0262] Constraint (23) indicates that the forklift needs to return to the turnover warehouse after completing the delivery task;

[0263] Constraint formula (24) indicates that the time it takes for the forklift to deliver the material to the workstation meets the time window requirement;

[0264] Expression (25) indicates the calculation method of the time it takes for the forklift to arrive at the delivery station;

[0265] Expression (26) indicates the earliest time of the hard time window required by workstation k to deliver material l time period;

[0266] Expression (27) indicates the latest time that material l arrives at workstation k;

[0267] Expression (28) indicates the safety stock of material l at workstation k;

[0268] The integrated model should take into account that during the packing process, the order of material packing should be opposite to the order of material unloading. Since the objectives of the various objective functions in distribution are different, the various objective functions are partially integrated through a multi-index comprehensive evaluation method. The indicators are dimensionless and the weights are determined. Finally, the integration of the workshop intelligent distributor model is completed. Thus, the above formulas (1) to (28) are integrated to obtain the calculation method of the total time loss of a distribution as shown in the following formula:

[0269]

[0270] The solution algorithm of the workshop intelligent distributor model based on Internet of Things technology is as follows: first establish four priority rules, list the information of the status of logistics and transportation equipment, and establish a list of logistics and transportation equipment and material distributors that can be called, a list of workstation requirements, a list of material distribution, and a list of what types of materials can be loaded in the turnover box to support the solution of the genetic algorithm, and then use the genetic algorithm as the main algorithm to solve the workshop intelligent distributor model.

[0271] The four priority rules are as follows:

[0272] 1. First, other constraints are considered only if the current remaining capacity of the logistics equipment can fully load all the required materials at the required delivery stations;

[0273] 2. Second, satisfy the time window constraints of the model to ensure that materials are delivered to the required workstations within the appropriate time period;

[0274] 3. Third, satisfy the shortest path constraint of the model. Logistics equipment delivers materials to the required workstations along the shortest path while meeting time constraints.

[0275] 4. Finally, if a workstation cannot meet its delivery needs in the current batch because its suitable delivery time period is significantly different from that of other workstations, or the current logistics and transportation equipment is not suitable for loading the required materials, its delivery needs will be automatically included in the next delivery batch.

[0276] The information on the status of the logistics transportation equipment is as follows:

[0277] (1) Assignment status, i.e. whether assigned;

[0278] (2) loading status, i.e. whether loading is completed;

[0279] (3) departure status, i.e., whether the delivery is being carried out;

[0280] (4) Delivery status, that is, whether the workstation requirements of the delivery list are met.

[0281] Among them, the genetic algorithm is used as the main algorithm to solve the workshop intelligent distributor model. The process of solving the workshop distributor model algorithm is as follows:

[0282] 1) Encoding and decoding: Based on multiple encoding methods, a multi-part encoding method is used to divide the gene into four parts. The encoded gene uses integers: the first part is the number of the material delivery person, the second part is the logistics transportation equipment number, the third part is the workstation delivery order, and the fourth part is the departure time of the material delivery person;

[0283] 2) Fitness function evaluation: the inverse of the objective function of the workshop intelligent distribution model is taken as its fitness function:

[0284]

[0285] 3) The selection in the basic operation adopts the roulette method. The calculation of the roulette in the algorithm is as follows:

[0286]

[0287]

[0288]

[0289]

[0290] Formula (a) is to calculate the fitness value of a single chromosome, formula (b) is to calculate the sum of the fitness values ​​of all individuals in the contemporary population, formula (c) is to calculate the selection probability of a single chromosome, and formula (d) is to calculate the cumulative probability. A random number r∈[0,1] is generated. If r∈[Q k-1 ,Q k ], then select individual x k , k∈(1,2,3…pop_size);

[0291] The crossover in the basic operation divides the chromosome into three parts. The first part is the delivery person number and the logistics transportation equipment number; the second part is the workstation delivery order, and the third part is the departure time of the material delivery person. The mutation operation is similarly divided into three parts. New offspring are obtained through the above operations, and finally the optimal complete logistics delivery plan is obtained through continuous selection and iteration.

[0292] Example 1:

[0293] This Example 1 is used as an example to verify the scientificity and effectiveness of the above-mentioned workshop intelligent dispenser model and its solution algorithm based on the Internet of Things technology:

[0294] refer to Figure 1 As shown, the workshop intelligent dispenser model based on Internet of Things technology and its solution algorithm of the present invention include the following steps:

[0295] (1) Real-time material information and workstation data are collected through IoT technology (RFID, workshop WLAN network, and smart application APP). For example, an assembly manufacturing workshop has six assembly lines. The current distribution task requires delivery to nine key workstations, and all materials are sourced from a single warehouse. The demand quantity and material demand time of each workstation are shown in Table 1, and the distance between workstations and travel time are shown in Table 2.

[0296] Table 1 Workstation information table

[0297]

[0298] Table 2 Distance between workstations and travel timetable

[0299]

[0300] Note: In Table 2, the area above the diagonal line is the travel time, and the area below the diagonal line is the distance between workstations.

[0301] (2) Set the basic parameters of the genetic algorithm, see Table 3

[0302] Table 3 Basic parameters

[0303]

[0304] The inverse of the objective function of the workshop intelligent distribution model is taken as its objective function:

[0305]

[0306] (3) Construct a multi-part coding method, divide the gene into four parts, and encode a delivery plan through this coding method.

[0307] (4) Randomly generate the initial population P within a reasonable value range, and the population size is selected as 100.

[0308] (5) Calculate the fitness function value of each individual in the population to evaluate the population.

[0309] (6) Select chromosomes in the population according to the rules and form a new population through crossover and mutation.

[0310] (7) Iteration: Set a suitable evolutionary generation. When the number of iterations is greater than the preset evolutionary generation or the convergence condition is reached, the optimization process ends.

[0311] (8) By comparing the results obtained by solving the workshop distributor model algorithm with those obtained by the classical genetic algorithm, it can be found that the workshop intelligent distributor algorithm is superior to the classical genetic algorithm in terms of optimization efficiency and convergence speed.

[0312] The present invention is specifically applied in the field of intelligent material distribution in manufacturing workshops. The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements can be made without departing from the principles of the present invention. These improvements should also be regarded as the scope of protection of the present invention.

Claims

1. A workshop intelligent distributor model based on Internet of Things technology, characterized by The modeling process is as follows: S1: Determine the decision goal By using the average delivery speed, packing load factor, vehicle route distance and total time loss during the delivery process as decision-making targets, the total time loss during delivery is reduced to a minimum, providing a systematically optimized delivery strategy for material delivery; Obtain the corresponding parameter information in the workshop logistics distribution, install readers and writers at the exit of the turnover warehouse, at each workstation of the assembly line and at the entrance of the finished product warehouse, and install electronic tags on the turnover boxes containing small materials, large materials and finished products, so as to realize the unique identification of workshop materials. The information of each material is contained in its electronic tag, and the RFID technology is used to quickly and accurately identify the materials; read the workstation data of the production site according to the Internet of Things equipment, mainly including the workstation number, the required material number, and the workstation data of the required material quantity, and then generate the workstation demand and enter it into the current demand list in the order of the latest delivery time of the materials; Among them, in the workshop logistics distribution, the corresponding parameters are obtained through the Internet of Things devices, and the parameter symbols are defined as follows: L——Length of turnover box W A ——Width of turnover box H——height of turnover box G——Maximum load capacity of turnover box N——Number of material types K——Number of workstations W——Number of forklifts G W ——Maximum load per vehicle P——Production line production rhythm, which should be measured according to the actual situation of the production line T min ——The minimum time that the current line inventory surplus can ensure the normal operation of the production line n——Logistics deliveryman m——Number of logistics equipment TS——Current delivery task number v i ——The average delivery speed of material delivery personnel (1≤i≤n, and i∈Z) v j ——The operating speed of logistics equipment (1≤j≤m, and j∈Z) v ij ——The ability of material delivery personnel i to work with logistics equipment j b i ——The maximum number of delivery tasks that deliveryman i can complete S ts ——The delivery path distance selected by the system for the current delivery task ts l——Material k——Workstation w——Forklift u——delivery order l l ——The length of the lth type of material (1≤l≤N, and l∈Z) w l ——Width of the lth type of material (1≤l≤N, and l∈Z) h l ——Height of the lth type of material (1≤l≤N, and l∈Z) g l ——Weight of the lth type of material (1≤l≤N, and l∈Z) q l ——The number of materials of the first category (1≤l≤N, and l∈Z) num l ——The number of materials of the first category loaded into the turnover box (1≤l≤N and l∈Z,0≤num l ≤q l And num l ∈Z) BWP l ——Height of the center of gravity of the lth type of material (1≤l≤N, and l∈Z) d kk' ——The distance from station k to station k' m k ——Total weight of materials required for the kth station BT kl ——The earliest time in the hard time window for the delivery of material l required by workstation k LT kl ——The latest time when material l arrives at workstation k RT kl ——The time when station k sends the material l demand instruction IL kl ——The inventory quantity of material l at the current line at workstation k TL kl ——The specific quantity of material l delivered to workstation k r kl ——Consumption rate of material l in workstation k SS kl ——Safety stock of material l at workstation k t kw ——The time it takes for forklift w to reach delivery station k [k]——the kth workstation in the delivery task list, k=0,1,2,…K; ——The time when material l arrives at the required workstation during the uth delivery by the material distributor ——The time when the material delivery person leaves the workstation where material l is located during the uth delivery. When l=0, this variable indicates the time when the material delivery person starts loading materials in the turnover warehouse. T S (l)——The loading time of material l in the turnover bin T u (l)——The unloading time of material l at the required station T kk’ (u)——The time it takes for the material delivery person to go from workstation k to workstation k' in the uth delivery, when k = 0 and k' = 0 Indicates that the material delivery person is in the turnover warehouse P k (t): Time loss function value of material delivery to workstation k at time t Constant λ(0≤λ≤1) Constants k1, k2, k3 (k1+k2+k3=1 and k1, k2, k3∈[0,1]); S2: The workshop intelligent dispatcher model must meet the following constraints:

1. Unify logistics containers and use turnover boxes of the same size during distribution; 2. Simplify and mark materials of different shapes and specifications into the smallest rectangular shape that meets their size; 3. The location of the turnover warehouse is determined and equipped with sufficient forklifts to distribute materials; 4. The materials in the turnover warehouse are sufficient and can meet the production needs of all workstations; 5. If the speed of all forklifts is determined and the same, the maximum distance that the forklift can travel is infinite; 6. Since the speed of forklifts is the same, if the distance traveled by the forklift is the shortest and the material demand of the workstation is met, the time spent on delivery is the shortest, that is, the total cost is the lowest; S3: Calculation formula and model integration in the process of establishing the workshop intelligent distributor model The calculation formula in the process of establishing the workshop intelligent distributor model is as follows: ST in ij =λv i +(1-λ)v j (10) F=k1CR+k2WR+k3CG (14) k1+k2+k3=1 and k1, k2, k3∈[0,1] (18) BT kl ≤t kw ≤LT kl (k=1,2,…,K) (24) In the above calculation formula, formula (1) represents the calculation formula of the comprehensive evaluation index of the multi-index comprehensive evaluation method; Expression (2) represents the weight calculation formula of the multi-index comprehensive evaluation method; Expression (3) represents the calculation formula of the standard deviation of the dimensionless standard value of each index; Constraint (4) restricts the range of weight values ​​in the calculation formula; Expressions (5) to (8) respectively represent the calculation formulas of the four indicators: the time consumed to complete the current delivery task, the space utilization rate, weight loading rate and overall center of gravity position of the comprehensive packaging, the length of the delivery path, and the total time loss of one delivery; Expression (9) represents the calculation method of the time consumed to complete the current delivery task; Expression (10) represents the calculation method of the working capacity of the current logistics personnel and logistics equipment; Constraint (11) indicates whether to assign a logistics delivery person to complete the task; Constraint (12) indicates that any delivery task can be completed by only one material delivery person; Constraint (13) indicates that the delivery tasks undertaken by the material delivery personnel shall not exceed the upper limit of their own work capacity; Expression (14) indicates the calculation method of the comprehensive container space utilization, weight loading rate and overall center of gravity position index; Expressions (15) to (17) respectively give the calculation methods of the functions expressing the packing space utilization, weight loading rate and overall center of gravity position; Constraint formula (18) limits the value range of the constant weight; Expression (19) indicates the calculation method of the delivery path length; Constraint formula (20) indicates that the total demand for materials at all workstations on each delivery route does not exceed the carrying capacity of the forklift; Constraint formula (21) indicates that the materials at each workstation are delivered by one and only one forklift; Constraint (22) indicates that each forklift needs to be dispatched from the turnover warehouse; Constraint (23) indicates that the forklift needs to return to the turnover warehouse after completing the delivery task; Constraint formula (24) indicates that the time it takes for the forklift to deliver the material to the workstation satisfies the time window requirement; Expression (25) indicates the calculation method of the time it takes for the forklift to arrive at the delivery station; Expression (26) indicates the earliest time of the hard time window required by workstation k to deliver material l time period; Expression (27) indicates the latest time that material l arrives at station k; Expression (28) indicates the safety stock of material l at workstation k; The integrated model should consider that during the packing process, the order of material packing should be opposite to the order of material unloading. Since the objectives of various objective functions in distribution are different, the various objective functions are integrated through the method of multi-index comprehensive evaluation. The indicators are dimensionless, the weights are determined, and finally the integration of the workshop intelligent distributor model is completed. Thus, the above formulas (1) to (28) are integrated to obtain the calculation method of the total time loss of a distribution as shown in the following formula: The solution algorithm of the model is as follows: firstly, four priority rules are set up, the information of the status of logistics and transportation equipment is listed, and the list of logistics and transportation equipment and material delivery personnel available for use, the list of workstation requirements, the list of material delivery, and the list of what types of materials can be loaded in the turnover box are established to support the solution of the genetic algorithm, and then the genetic algorithm is used as the main algorithm to solve the workshop intelligent distributor model; In the solution of the workshop intelligent distributor model using genetic algorithm as the main algorithm, the process of solving the workshop distributor model algorithm is as follows: 1) Encoding and decoding, based on multiple encoding methods, a multi-part encoding method is used to divide the gene into four parts for encoding. The encoded gene uses integers: the first part is the number of the material delivery person, the second part is the logistics transportation equipment number, the third part is the station distribution sequence, and the fourth part is the departure time of the material delivery person; 2) Fitness function evaluation: the inverse of the objective function of the workshop intelligent distribution model is taken as its fitness function: 3) The selection in the basic operation adopts the roulette method, and the calculation of the roulette in the algorithm is as follows: Formula (a) is to calculate the fitness value of a single chromosome, formula (b) is to calculate the sum of the fitness values ​​of all individuals in the contemporary population, formula (c) is to calculate the selection probability of a single chromosome, and formula (d) is to calculate the cumulative probability. The random number r∈[0,1] is generated. If r∈[Q k-1 ,Q k ], then select individual x k , k∈(1,2,3…pop_size); The crossover in the basic operation divides the chromosome into three parts. The first part is the delivery staff number and the logistics transportation equipment number; the second part is the workstation delivery order, and the third part is the departure time of the material delivery staff. The mutation operation is similarly divided into three parts. New offspring are obtained through the above operations, and finally the optimal complete logistics distribution plan is obtained through continuous selection and iteration.

2. A workshop intelligent distributor model based on Internet of Things technology as claimed in claim 1, characterized in that The four priority rules are as follows:

1. First, other constraints are considered only if and only if the current remaining capacity of the logistics equipment can fully load all the required materials at the required delivery stations; 2. Second, meet the time window constraints of the model to ensure that the materials are delivered to the required workstations within the appropriate time period; 3. Third, satisfy the shortest path constraint of the model. On the premise of meeting the time constraint, the logistics equipment delivers the materials to the required workstation along the shortest path; 4. Finally, if a certain workstation cannot meet its distribution needs in the current batch because the time period suitable for distribution is significantly different from the time period required by other workstations, or the current logistics and transportation equipment is not suitable for loading the required materials, its distribution needs will be automatically included in the next distribution batch.

3. A workshop intelligent distributor model based on Internet of Things technology as claimed in claim 1, characterized in that The information on the status of the logistics transportation equipment is as follows: (1) Assignment status, i.e. whether it is assigned; (2) loading status, i.e. whether loading is completed; (3) departure status, i.e., whether the delivery is being carried out; (4) Delivery status, that is, whether the workstation requirements in the delivery list are met.

Images