Automatic loading method and system for grain warehouse

Abstract

The invention discloses an automatic loading method and system for a grain warehouse, and the method comprises the steps: collecting the operation data and loading orders of the grain warehouse, dividing the loading process of the grain warehouse into simple processes, and constructing a topological network model; predicting the operation data to obtain predicted operation data, and obtaining a congestion risk index of the topology network model based on the predicted operation data and the operation data; training an intelligent planning model based on the topological network model according to the operation data and the congestion risk index, and establishing a granary automatic loading model based on the intelligent planning model; and inputting a to-be-planned order into the granary automatic loading model to obtain a planned path. The method not only can improve the automatic loading precision of the grain warehouse, but also has good interpretability, and can be directly applied to an automatic loading system of the grain warehouse.

Description

Technical Field

[0001] The present invention relates to the field of automation technology, and in particular to an automatic loading method and system for a grain warehouse. Background Art

[0002] With the rapid development of modern logistics industry, grain warehouses are key nodes in the logistics system, and their loading efficiency and accuracy are crucial to the operation of the entire supply chain. Traditional grain warehouse loading methods mainly rely on manual operations, which have problems such as low efficiency, high error rate, and high labor intensity. In addition, due to the lack of effective automation and intelligent means, grain warehouses often find it difficult to respond quickly to market changes when facing large-scale and highly dynamic loading needs, resulting in waste of resources and increased costs. In order to solve these problems, in recent years, automation technology and intelligent algorithms have been introduced into grain warehouse management, in order to improve loading efficiency and accuracy, reduce operating costs, and enhance the flexibility and adaptability of the system through technical means. However, most of the existing automated loading systems are limited to simple process control, lack in-depth understanding and optimization of complex loading processes, and are difficult to cope with changing loading environments and needs. Therefore, it has become an urgent need for the development of the industry to develop a grain warehouse automatic loading method and system that can adapt to complex loading environments, with high efficiency and high accuracy. Summary of the invention

[0003] The object of the present invention is to provide a method and system for automatically loading grain in a grain warehouse.

[0004] To achieve the above object, the present invention is implemented according to the following technical solutions: The present invention comprises the following steps: Collect the operation data and loading orders of the grain warehouse, divide the loading process of the grain warehouse into simple processes, and build a topological network model; Predicting the operation data to obtain predicted operation data, and obtaining a congestion risk index of the topological network model based on the predicted operation data and the operation data; According to the operation data and the congestion risk index, an intelligent planning model is trained based on the topological network model, and an automatic loading model for a granary is established based on the intelligent planning model; The order to be planned is input into the grain silo automatic loading model to obtain a planned path.

[0005] Furthermore, the loading process of the grain warehouse is divided into simple processes, and the method of constructing a topological network model includes: The loading process includes the complete transportation process of goods from the starting point to the end point. The loading process is converted into a directed graph, with warehouse transfer points as nodes and simple processes as edges. The ratio of the capacity of the warehouse transfer point to the total capacity is the initial weight of the node, and the ratio of the length of the simple process to the total length is the initial weight of the edge. Node and edge weights: in, For Node The weight of For edge The weight of and are the influence weights of node capacity and processing speed respectively, and are the influence weights of edge length and transportation time, respectively. and Node capacity and processing speed, and Side length and normal transportation time, , , and They are , , and The maximum value of .

[0006] Furthermore, a method for predicting the operation data to obtain predicted operation data includes: The prediction model is trained using running data, and the quantile objective function of the prediction model is: in, for The q-quantile objective function under is the quantile, , represents the set of model parameters of the prediction model, The prediction model model parameters, is the number of categories of running data, is the sequence length of the running data, For the time step The e-th type of operation data, is the e-th type of running data in the prediction model at time step The forecast data, is the coefficient of the L1 regularization term, For the allowable error, take , is an indicator function, if If yes, it is 1, otherwise it is 0. is the loss function of the prediction model, Forget Gate: in, is the output of the forget gate, is the sigmoid activation function, H is the dimension of the hidden layer, is the attention weight vector, is the initial weight of the jth dimension of the weight matrix in the forget gate, for The learning rate at the current time step, express and The jth dimension of is the hidden state at the previous time step, is the independent variable of the current time step, b is the bias term, Output Gate: in, is the output of the current output gate, is the j-th dimension output of the output gate, is the j-th dimension bias term of the cell state, is the hidden state at the current time step, is the initial weight of the j-th dimension in the output gate, for The learning rate at the current time step, Output the forecast data sequence 4 hours after each running data as the forecast running data.

[0007] Furthermore, the method for obtaining the congestion risk index of the topological network model based on the predicted operation data and the operation data includes: Fixed risk of nodes and edges: in, For Node The fixed risk , , , and are the influence weights of clustering coefficient, betweenness, degree, degree distribution and initial weight of nodes, respectively. , , and are the clustering coefficient, betweenness, degree and degree distribution of node i respectively; For edge The congestion risk index, , and are the impact weights of transportation time, transportation capacity and adjacent node congestion risk index, , , and Side The transportation time, transportation capacity, congestion risk index of adjacent node 1 and fixed risk of adjacent node 2; Calculate the congestion risk index of nodes and edges based on the predicted operation data, operation data and fixed risk: in, and are the current time step nodes respectively and edge The congestion risk index, and are the current time step nodes respectively and edge of traffic, and are the current time step nodes respectively and edge The average value of the predicted flow, and are the current time step nodes respectively and edge The processing time of the current traffic, and are the current time step nodes respectively and edge The average value of the predicted processing time for the current traffic, and For Node and edge The congestion threshold for the processing time.

[0008] Furthermore, the method for training an intelligent planning model based on the topological network model according to the operation data and the congestion risk index includes: Reinforcement learning is used to train the intelligent planning model. The environment is a topological network model. The intelligent agents include ordered grain and ordered vehicles. The state includes the location of ordered grain and ordered vehicles. The environment includes the real-time inventory of grain in the warehouse, the weights of nodes and edges, and the congestion risk index. The action space of ordered grain includes selecting warehouses and selecting transportation routes. The action space of ordered grain is to select loading spaces. Rewards are composed of loading time, order completion accuracy, congestion risk index of transportation routes, equipment energy consumption, and vehicle energy consumption. Update of value function: in is the value function of state S, is the weight of action A, is the strategy for choosing action A in state S, is the immediate reward for executing action A, α is the discount factor of the value function, which is used to adjust the current value of future rewards, It is to execute action A from state S to state The probability of Update of Q function: in is the Q function of state S and action A, is the expected estimate of taking action A in state S, β is the discount factor of the Q function, To move from state S to state S, take action A and get rewarded The probability of is the state when the discount factor of the Q function is used The value function of is the parameter of the main network used to parameterize the Q function, E(·) represents the expectation, which represents the average value under the probability distribution, is an adaptive weight that is dynamically adjusted based on the historical performance of action A in state S, renew The policy gradient formula is: in is the objective function The gradient of is the expected return, is an adaptive weight that is dynamically adjusted based on the historical performance of action B in state S, For the status The immediate reward obtained when executing action B, B is the currently selected action, To transition from state S to state B The probability of represents the discount factor of the policy gradient, Status The value function of It is a parameterized strategy that is iteratively optimized until the strategy converges.

[0009] Furthermore, a method for establishing a grain silo automatic loading model based on the intelligent planning model includes: The intelligent planning model is used as the basic model of the grain warehouse automatic loading model, and the basic model is used as the starting point of incremental learning. When the grain types and equipment facilities in the grain warehouse change, the patterns and rules in the new data are quickly learned based on the basic model.

[0010] An automatic loading system for a grain warehouse, characterized by comprising: The collection module collects the operation data and loading orders of grain warehouses; A calculation module divides the loading process of the grain warehouse into simple processes, constructs a topological network model, predicts the operation data to obtain predicted operation data, and obtains a congestion risk index of the topological network model based on the predicted operation data and the operation data; An intelligent planning module, which trains an intelligent planning model based on the topological network model according to the operation data and the congestion risk index, and establishes an automatic loading model for a granary based on the intelligent planning model; The output module inputs the order to be planned into the grain silo automatic loading model to obtain the planned path.

[0011] In a second aspect, an embodiment of the present application further provides an electronic device, including: A processor; and a memory arranged to store computer executable instructions, which when executed cause the processor to perform the method steps described in the first aspect.

[0012] In a third aspect, an embodiment of the present application further provides a computer-readable storage medium, which stores one or more programs. When the one or more programs are executed by an electronic device including multiple applications, the electronic device executes the method steps described in the first aspect.

[0013] The beneficial effects of the present invention are: The present invention significantly improves the accuracy and efficiency of automatic loading of grain warehouses by constructing a topological network model, predicting operation data, calculating the congestion risk index, and combining reinforcement learning to train an intelligent planning model. The system not only reduces manual intervention and labor intensity, but also enhances the adaptability to complex loading environments, effectively reduces operating costs, and improves the intelligence level of loading operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1A flowchart of the steps of an automatic loading method and system for a grain warehouse according to the present invention; Figure 2 It is a schematic diagram of the structure of an electronic device in an embodiment of this specification. DETAILED DESCRIPTION

[0015] The present invention is further described below by means of specific embodiments. The illustrative embodiments and descriptions of the present invention are used to explain the present invention but are not intended to limit the present invention.

[0016] The present invention provides a method and system for automatically loading a grain warehouse, comprising the following steps: like Figure 1 As shown, in this embodiment, the following steps are included: Collect the operation data and loading orders of the grain warehouse, divide the loading process of the grain warehouse into simple processes, and build a topological network model; In the actual evaluation, a medium-sized grain warehouse was selected. The warehouse is equipped with various sensors, including grain inventory sensors, transportation equipment operation status sensors and vehicle positioning sensors. At the same time, multiple high-definition cameras are installed for video monitoring. The internal layout of the warehouse is relatively complex, including multiple transfer points and different types of transportation channels, storing wheat, corn and rice. The loading and unloading equipment includes conveyor belts and loaders. The transportation vehicles include trucks with different load specifications. The operation data includes the operation log data of the past year, including the daily storage location and storage volume changes of various types of grain, the detailed time record and actual loading path of each historical loading operation, the layout diagram of the warehouse, and the width, length and carrying capacity of the channel. The loading order includes the order number, grain type and quantity. The loading order and the operation data are matched according to time. The nodes of the topological network model include the grain warehouse entrance, grain storage point, loading parking space, conveyor belt and transfer station. The simple process connecting the nodes is the edge of the topological network model. Predicting the operation data to obtain predicted operation data, and obtaining a congestion risk index of the topological network model based on the predicted operation data and the operation data; In the actual evaluation, the clustering coefficient of node A is 0.3, the betweenness is 0.2, the degree is 5, the degree distribution is 0.1, the initial weight of the node is 0.2, and the influence weights of the clustering coefficient, betweenness, degree, degree distribution and initial weight of the node are 0.1, 0.2, 0.3, 0.1 and 0.3 respectively. The fixed risk of the node is calculated to be 1.1969 by the formula. At time C, the current flow of node A is 30 tons / hour, the predicted average flow is 35 tons / hour, the processing time is 6 minutes, the predicted average processing time is 5 minutes, and the congestion threshold of the processing time is 8 minutes. According to the formula, the congestion risk index of the node at this time is 1.375. According to the operation data and the congestion risk index, an intelligent planning model is trained based on the topological network model, and an automatic loading model for a granary is established based on the intelligent planning model; In the actual evaluation, in the reward function of the intelligent planning model, the weight of loading time is -0.4, the weight of order completion accuracy is 0.6, the weight of congestion risk index is -0.1, the weight of equipment energy consumption is -0.1, and the weight of vehicle energy consumption is -0.1; The order to be planned is input into the grain silo automatic loading model to obtain a planned path.

[0017] In this embodiment, the loading process of the grain warehouse is divided into simple processes, and the method for constructing a topological network model includes: The loading process includes the complete transportation process of goods from the starting point to the end point. The loading process is converted into a directed graph, with warehouse transfer points as nodes and simple processes as edges. The ratio of the capacity of the warehouse transfer point to the total capacity is the initial weight of the node, and the ratio of the length of the simple process to the total length is the initial weight of the edge. Node and edge weights: in, For Node The weight of For edge The weight of and are the influence weights of node capacity and processing speed respectively, and are the influence weights of edge length and transportation time, respectively. and Node capacity and processing speed, and Side length and normal transportation time, , , and They are , , and The maximum value of .

[0018] In this embodiment, the method of predicting the operation data to obtain predicted operation data includes: The prediction model is trained using running data, and the quantile objective function of the prediction model is: in, for The q-quantile objective function under is the quantile, , represents the set of model parameters of the prediction model, The prediction model model parameters, is the number of categories of running data, is the sequence length of the running data, For the time step The e-th type of operation data, is the e-th type of running data in the prediction model at time step The forecast data, is the coefficient of the L1 regularization term, For the allowable error, take , is an indicator function, if If yes, it is 1, otherwise it is 0. is the loss function of the prediction model, Forget Gate: in, is the output of the forget gate, is the sigmoid activation function, H is the dimension of the hidden layer, is the attention weight vector, is the initial weight of the jth dimension of the weight matrix in the forget gate, for The learning rate at the current time step, express and The jth dimension of is the hidden state at the previous time step, is the independent variable of the current time step, b is the bias term, Output Gate: in, is the output of the current output gate, is the j-th dimension output of the output gate, is the j-th dimension bias term of the cell state, is the hidden state at the current time step, is the initial weight of the j-th dimension in the output gate, for The learning rate at the current time step, Output the forecast data sequence 4 hours after each running data as the forecast running data.

[0019] In this embodiment, the method for obtaining the congestion risk index of the topological network model based on the predicted operation data and the operation data includes: Fixed risk of nodes and edges: in, For Node The fixed risk , , , and are the influence weights of clustering coefficient, betweenness, degree, degree distribution and initial weight of nodes, respectively. , , and are the clustering coefficient, betweenness, degree and degree distribution of node i respectively; For edge The congestion risk index, , and are the impact weights of transportation time, transportation capacity and adjacent node congestion risk index, , , and Side The transportation time, transportation capacity, congestion risk index of adjacent node 1 and fixed risk of adjacent node 2; Calculate the congestion risk index of nodes and edges based on the predicted operation data, operation data and fixed risk: in, and are the current time step nodes respectively and edge The congestion risk index, and are the current time step nodes respectively and edge of traffic, and are the current time step nodes respectively and edge The average value of the predicted flow, and are the current time step nodes respectively and edge The processing time of the current traffic, and are the current time step nodes respectively and edge The average value of the predicted processing time for the current traffic, and For Node and edge The congestion threshold for the processing time.

[0020] In this embodiment, the method for training an intelligent planning model based on the topological network model according to the operation data and the congestion risk index includes: Reinforcement learning is used to train the intelligent planning model. The environment is a topological network model. The intelligent agents include ordered grain and ordered vehicles. The state includes the location of ordered grain and ordered vehicles. The environment includes the real-time inventory of grain in the warehouse, the weights of nodes and edges, and the congestion risk index. The action space of ordered grain includes selecting warehouses and selecting transportation routes. The action space of ordered grain is to select loading spaces. The loading time, order completion accuracy, congestion risk index of the transportation path, equipment energy consumption, and vehicle energy consumption are weighted and summed according to positive impact, negative impact, and impact importance to obtain rewards. Update of value function: in is the value function of state S, is the weight of action A, is the strategy for choosing action A in state S, is the immediate reward for executing action A, α is the discount factor of the value function, which is used to adjust the current value of future rewards, It is to execute action A from state S to state The probability of Update of Q function: in is the Q function of state S and action A, is the expected estimate of taking action A in state S, β is the discount factor of the Q function, To move from state S to state S, take action A and get rewarded The probability of is the state when the discount factor of the Q function is used The value function of is the parameter of the main network used to parameterize the Q function, E(·) represents the expectation, which represents the average value under the probability distribution, is an adaptive weight that is dynamically adjusted based on the historical performance of action A in state S, renew The policy gradient formula is: in is the objective function The gradient of Express Derived, for the expected return, is an adaptive weight that is dynamically adjusted based on the historical performance of action B in state S, For the status The instant reward obtained when executing action B is the instant reward of the current action, and B is the currently selected action. To transition from state S to state B The probability of is the transition probability of the future state, represents the discount factor of the policy gradient, Status The value function is the value of the future state, It is a parameterized strategy that is iteratively optimized until the strategy converges.

[0021] In this embodiment, a method for establishing a grain silo automatic loading model based on the intelligent planning model includes: The intelligent planning model is used as the basic model of the grain warehouse automatic loading model, and the basic model is used as the starting point of incremental learning. When the grain types and equipment facilities in the grain warehouse change, the patterns and rules in the new data are quickly learned based on the basic model.

[0022] An automatic loading system for a grain warehouse, characterized by comprising: The collection module collects the operation data and loading orders of grain warehouses; A calculation module divides the loading process of the grain warehouse into simple processes, constructs a topological network model, predicts the operation data to obtain predicted operation data, and obtains a congestion risk index of the topological network model based on the predicted operation data and the operation data; An intelligent planning module, which trains an intelligent planning model based on the topological network model according to the operation data and the congestion risk index, and establishes an automatic loading model for a granary based on the intelligent planning model; The output module inputs the order to be planned into the grain silo automatic loading model to obtain the planned path.

[0023] Figure 2 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present application. Figure 2At the hardware level, the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and a memory. The memory may include a memory, such as a high-speed random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk storage. Of course, the electronic device may also include hardware required for other services.

[0024] The processor, network interface and memory can be interconnected through an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 2 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0025] The memory is used to store the program. Specifically, the program may include a program code, and the program code includes a computer operation instruction. The memory may include a memory and a non-volatile memory, and provides instructions and data to the processor.

[0026] The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it, forming an automatic loading device for grain warehouses at the logical level. The processor executes the program stored in the memory and is specifically used to execute any of the aforementioned automatic loading methods and systems for grain warehouses.

[0027] The above application Figure 1The automatic loading method and system for grain warehouse disclosed in the illustrated embodiment can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capability. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or an instruction in software form. The above processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiments of the present application can be directly embodied as being executed by a hardware decoding processor, or being executed by a combination of hardware and software modules in a decoding processor. The software module can be located in a storage medium mature in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.

[0028] The electronic device may also perform Figure 1 Automatic loading method and system for grain warehouse in China, and realize Figure 1 The functions of the illustrated embodiment will not be described in detail in the embodiments of the present application.

[0029] An embodiment of the present application also proposes a computer-readable storage medium, which stores one or more programs, and the one or more programs include instructions. When the instructions are executed by an electronic device including multiple application programs, they execute any of the aforementioned automatic loading methods and systems for grain warehouses.

[0030] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0031] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0032] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0033] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0034] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0035] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0036] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0037] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0038] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0039] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for automatic loading of grain warehouse, characterized in that: The following steps are involved: Collect the operation data and loading orders of the grain warehouse, divide the loading process of the grain warehouse into simple processes, and build a topological network model; Predicting the operation data to obtain predicted operation data, and obtaining a congestion risk index of the topological network model based on the predicted operation data and the operation data; According to the operation data and the congestion risk index, an intelligent planning model is trained based on the topological network model, and an automatic loading model for a granary is established based on the intelligent planning model; The order to be planned is input into the grain silo automatic loading model to obtain a planned path.

2. According to claim 1, a method for automatically loading a grain warehouse is characterized in that: The method of dividing the loading process of the grain warehouse into simple processes and constructing a topological network model includes: The loading process includes the complete transportation process of goods from the starting point to the end point. The loading process is converted into a directed graph, with warehouse transfer points as nodes and simple processes as edges. The ratio of the capacity of the warehouse transfer point to the total capacity is the initial weight of the node, and the ratio of the length of the simple process to the total length is the initial weight of the edge. Node and edge weights: in, For Node The weight of For edge The weight of and are the influence weights of node capacity and processing speed respectively, and are the influence weights of edge length and transportation time, respectively. and Node capacity and processing speed, and Side length and normal transportation time, , , and They are , , and The maximum value of .

3. According to claim 1, a method for automatically loading a grain warehouse is characterized in that: The method for predicting the operation data to obtain predicted operation data comprises: The prediction model is trained using running data, and the quantile objective function of the prediction model is: in, for The q-quantile objective function under is the quantile, , represents the set of model parameters of the prediction model, The prediction model model parameters, is the number of categories of running data, is the sequence length of the running data, For the time step The e-th type of operation data, is the e-th type of running data in the prediction model at time step The forecast data, is the coefficient of the L1 regularization term, For the allowable error, take , is an indicator function, if If yes, it is 1, otherwise it is 0. is the loss function of the prediction model, Forget Gate: in, is the output of the forget gate, is the sigmoid activation function, H is the dimension of the hidden layer, is the attention weight vector, is the initial weight of the jth dimension of the weight matrix in the forget gate, for The learning rate at the current time step, express and The jth dimension of is the hidden state at the previous time step, is the independent variable of the current time step, b is the bias term, Output Gate: in, is the output of the current output gate, is the j-th dimension output of the output gate, is the j-th dimension bias term of the cell state, is the hidden state at the current time step, is the initial weight of the j-th dimension in the output gate, for The learning rate at the current time step, Output the forecast data sequence 4 hours after each running data as the forecast running data.

4. According to claim 1, a method for automatically loading a grain warehouse is characterized in that: The method for obtaining the congestion risk index of the topological network model based on the predicted operation data and the operation data includes: Fixed risk of nodes and edges: in, For Node The fixed risk , , , and are the influence weights of clustering coefficient, betweenness, degree, degree distribution and initial weight of nodes, respectively. , , and are the clustering coefficient, betweenness, degree and degree distribution of node i respectively; For edge The congestion risk index, , and are the impact weights of transportation time, transportation capacity and adjacent node congestion risk index, , , and Side The transportation time, transportation capacity, congestion risk index of adjacent node 1 and fixed risk of adjacent node 2; Calculate the congestion risk index of nodes and edges based on the predicted operation data, operation data and fixed risk: in, and are the current time step nodes respectively and edge The congestion risk index, and are the current time step nodes respectively and edge of traffic, and are the current time step nodes respectively and edge The average value of the predicted flow, and are the current time step nodes respectively and edge The processing time for the current traffic, and are the current time step nodes respectively and edge The average value of the predicted processing time for the current traffic, and For Node and edge The congestion threshold for the processing time.

5. The automatic loading method for grain warehouse according to claim 1 is characterized in that: The method for training an intelligent planning model based on the topological network model according to the operation data and the congestion risk index includes: Reinforcement learning is used to train the intelligent planning model. The environment is a topological network model. The intelligent agents include ordered grain and ordered vehicles. The state includes the location of ordered grain and ordered vehicles. The environment includes the real-time inventory of grain in the warehouse, the weights of nodes and edges, and the congestion risk index. The action space of ordered grain includes selecting warehouses and selecting transportation routes. The action space of ordered grain is to select loading spaces. Rewards are composed of loading time, order completion accuracy, congestion risk index of transportation routes, equipment energy consumption, and vehicle energy consumption. Update of value function: in is the value function of state S, is the weight of action A, is the strategy for choosing action A in state S, is the immediate reward for executing action A, α is the discount factor of the value function, which is used to adjust the current value of future rewards, It is to execute action A from state S to state The probability of Update of Q function: in is the Q function of state S and action A, is the expected estimate of taking action A in state S, β is the discount factor of the Q function, To move from state S to state S, take action A And get rewards The probability of is the state when the discount factor of the Q function is used The value function of is the parameter of the main network used to parameterize the Q function, E(·) represents the expectation, which represents the average value under the probability distribution, is an adaptive weight that is dynamically adjusted based on the historical performance of action A in state S, renew The policy gradient formula is: in is the objective function The gradient of is the expected return, is an adaptive weight that is dynamically adjusted based on the historical performance of action B in state S, For the status The immediate reward obtained when executing action B, B is the currently selected action, To transition from state S to state B The probability of represents the discount factor of the policy gradient, Status The value function of It is a parameterized strategy that is iteratively optimized until the strategy converges.

6. The automatic loading method for grain warehouse according to claim 1, characterized in that: The method for establishing a grain silo automatic loading model based on the intelligent planning model includes: The intelligent planning model is used as the basic model of the grain warehouse automatic loading model, and the basic model is used as the starting point of incremental learning. When the grain types and equipment facilities in the grain warehouse change, the patterns and rules in the new data are quickly learned based on the basic model.

7. An automatic loading system for grain warehouses, used to execute the automatic loading method for grain warehouses according to any one of claims 1 to 6, characterized in that: include: The collection module collects the operation data and loading orders of grain warehouses; A calculation module divides the loading process of the grain warehouse into simple processes, constructs a topological network model, predicts the operation data to obtain predicted operation data, and obtains a congestion risk index of the topological network model based on the predicted operation data and the operation data; An intelligent planning module, which trains an intelligent planning model based on the topological network model according to the operation data and the congestion risk index, and establishes an automatic loading model for a granary based on the intelligent planning model; The output module inputs the order to be planned into the grain silo automatic loading model to obtain the planned path.

8. An electronic device comprising: processor; as well as A memory arranged to store computer executable instructions, which when executed cause the processor to perform the method of any one of claims 1 to 6.

9. A computer-readable storage medium storing one or more programs, which, when executed by an electronic device including a plurality of application programs, enables the electronic device to execute any one of the methods of claims 1 to 6.

Images