Intelligent path planning method in container loading and unloading process

Abstract

The invention provides an intelligent path planning method in a container loading and unloading process, and relates to the technical field of loading and unloading path optimization, and the method comprises the steps: distributing a plurality of slide rail nodes, and constructing a first slide rail topology network; optimizing the first slide rail topology network by using a plurality of loading and unloading nodes, and constructing a second slide rail topology network; constructing a path planning model based on the second slide rail topology network; and inputting the real-time loading and unloading task information, the sliding rail real-time monitoring information and the environment real-time monitoring information into a path planning model for path planning, and outputting a loading and unloading planning path. Through the method, the technical problems that a dynamic change environment cannot be quickly responded, loading and unloading path conflicts, resource waste and the like are caused, and loading and unloading efficiency and quality are affected due to the fact that a traditional path planning method is difficult to balance a plurality of optimization targets can be solved, the complex decision problem of multi-target optimization can be efficiently processed, and loading and unloading efficiency and quality are improved. The scientificity and accuracy of loading and unloading path setting are improved, and the loading and unloading efficiency, quality and safety are effectively improved.

Description

Technical Field

[0001] This application relates to the field of loading and unloading route optimization technology, and in particular to an intelligent route planning method for container loading and unloading processes. Background Technology

[0002] In modern ports and logistics centers, container handling is a highly complex and dynamic process. With the continuous increase in global trade volume, the demand for container processing is growing, and the efficiency and quality of handling operations directly impact the operation of the entire supply chain. However, the container handling environment is typically highly complex and uncertain, and traditional route planning methods have significant limitations in handling these dynamic changes, making it difficult to achieve optimal route planning solutions in such volatile environments.

[0003] Currently, due to the complex and ever-changing container loading and unloading environment, traditional route planning methods are unable to balance multiple optimization objectives, resulting in an inability to quickly respond to dynamically changing environments and accurately set appropriate loading and unloading routes. This leads to phenomena such as loading and unloading route conflicts and resource waste, which affect the efficiency and quality of container loading and unloading. Summary of the Invention

[0004] The purpose of this application is to provide an intelligent path planning method for container loading and unloading, in order to solve the technical problems that traditional path planning methods are unable to balance multiple optimization objectives due to the complex and ever-changing container loading and unloading environment, resulting in the inability to quickly respond to dynamic changes and accurately set appropriate loading and unloading paths, causing loading and unloading path conflicts, resource waste, and other phenomena that affect the efficiency and quality of container loading and unloading.

[0005] In view of the above problems, this application provides an intelligent path planning method for container loading and unloading, comprising: distributing multiple sliding rail nodes in the container loading and unloading area, the multiple sliding rail nodes being connected by sliding rails to form a path map, and constructing a first sliding rail topology network, wherein the first sliding rail topology network is controlled by a sliding rail system; determining multiple loading and unloading nodes in the container loading and unloading area, the multiple loading and unloading nodes including loading and unloading platform nodes, container yard nodes, and forklift loading and unloading nodes; optimizing the first sliding rail topology network with the multiple loading and unloading nodes to construct a second sliding rail topology network; constructing a path planning model based on the second sliding rail topology network; the sliding rail system receiving real-time loading and unloading task information, real-time sliding rail monitoring information, and real-time environmental monitoring information, inputting the real-time loading and unloading task information, the real-time sliding rail monitoring information, and the real-time environmental monitoring information into the path planning model for path planning, and outputting the loading and unloading planned path.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages: By distributing multiple sliding rail nodes in the container loading and unloading area, and connecting these nodes with sliding rails to form a path graph, a first sliding rail topology network is constructed. Next, multiple loading and unloading nodes in the container loading and unloading area are determined, and the first sliding rail topology network is optimized using these nodes to construct a second sliding rail topology network. Then, a path planning model is built based on the second sliding rail topology network. Finally, real-time loading and unloading task information, real-time sliding rail monitoring information, and real-time environmental monitoring information are received by the sliding rail system and input into the path planning model for path planning, outputting the planned loading and unloading path. This method can quickly adapt to the dynamically changing environment during container loading and unloading, efficiently handle complex decision-making problems involving multi-objective optimization, thereby improving the scientific nature and accuracy of container loading and unloading path setting, and effectively enhancing the efficiency, quality, and safety of container loading and unloading.

[0007] The above description is merely an overview of the technical solution of this application. To better understand the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0009] Figure 1 This is a flowchart illustrating an intelligent path planning method for container loading and unloading according to this application. Figure 2 This is a schematic diagram of the process of updating the first slide rail topology network in the intelligent path planning method for container loading and unloading in this application. Detailed Implementation

[0010] This application provides an intelligent path planning method for container loading and unloading, solving the technical problem that traditional path planning methods struggle to balance multiple optimization objectives due to the complex and ever-changing container loading and unloading environment. This leads to an inability to quickly respond to dynamically changing environments and accurately set appropriate loading and unloading paths, resulting in path conflicts, resource waste, and other issues that negatively impact container loading and unloading efficiency and quality. The method described above can quickly adapt to the dynamically changing environment during container loading and unloading, efficiently handle complex multi-objective optimization decision-making problems, thereby improving the scientific rigor and accuracy of container loading and unloading path setting, and effectively enhancing the efficiency, quality, and safety of container loading and unloading.

[0011] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. It should be understood that this application is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application. It should also be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all of them.

[0012] For examples, please refer to the appendix. Figure 1 This application provides an intelligent path planning method for container loading and unloading processes, specifically including the following steps: Step 1: Distribute multiple slide rail nodes in the container loading and unloading area. The multiple slide rail nodes are connected by slide rails to form a path map, thus constructing a first slide rail topology network. The first slide rail topology network is controlled by a slide rail system.

[0013] Specifically, within the container loading and unloading area, multiple sliding rail nodes are distributed according to actual operational needs and site layout. The rational distribution of these nodes covers the entire loading and unloading area, enabling the trolleys to move flexibly between various important operational points to complete container transportation tasks. These multiple sliding rail nodes are connected by rails to form a path map. By designing reasonable routes and connection methods, the smooth movement of the trolleys between nodes is ensured. The sliding rail path map formed by connecting the sliding rail nodes constitutes the first sliding rail topology network. This topology network not only reflects the physical connections but also demonstrates the path layout of the entire sliding rail system through its topological structure. In this network, the sliding rail nodes act as network nodes, and the rails act as network edges, forming a complete graph structure. The first sliding rail topology network is controlled by the sliding rail system, an intelligent management system composed of a central control unit, a sensor network, and a communication network. This system enables path planning and scheduling management of the trolleys, ensuring efficiency and accuracy during container loading and unloading.

[0014] Step 2: Identify multiple loading and unloading nodes in the container loading and unloading area, including loading and unloading platform nodes, container yard nodes, and forklift loading and unloading nodes.

[0015] Specifically, within the container loading and unloading area, several key loading and unloading nodes are identified. These nodes are the main locations where the trolleys in the sliding rail system perform container loading and unloading tasks. These nodes include loading and unloading platform nodes, container yard nodes, and forklift loading and unloading nodes. Loading and unloading platform nodes are typically located near the container loading and unloading platform and are the main locations for loading and unloading containers from ships or other means of transport. Trolleys receive or place containers at these nodes and then transport them to designated locations. Container yard nodes are distributed within the container yard and serve as temporary storage locations for containers. At these nodes, trolleys can stack containers in designated locations within the yard or retrieve containers from the yard for further transport or loading and unloading operations. Forklift loading and unloading nodes are located in the forklift operation area and are typically used to load containers from trolleys onto forklifts, or to load containers transported by forklifts onto trolleys. Forklift loading and unloading nodes are crucial for transferring containers between different modes of transport. By identifying these key loading and unloading nodes, a comprehensive loading and unloading path network can be constructed, ensuring that trolleys can move efficiently between different nodes to complete various loading and unloading tasks.

[0016] Step 3: Optimize the first slide rail topology network using the multiple loading and unloading nodes to construct a second slide rail topology network.

[0017] Specifically, the first slide rail topology network is optimized using the multiple loading and unloading nodes to construct a second slide rail topology network. First, the identified loading and unloading nodes (including loading and unloading platform nodes, container yard nodes, and forklift loading and unloading nodes) are analyzed to assess the importance of each node in the entire container loading and unloading process. This can be determined based on factors such as the node's task frequency, load conditions, and connectivity with other nodes. Next, based on the importance of the loading and unloading nodes, the first slide rail topology network is adjusted. For example, adding new slide rails or adjusting the connection method of existing slide rails can enhance the connectivity between important nodes, shorten the movement path of the trolley between these key nodes, and thus reduce transportation time and energy consumption. Bottleneck nodes in the first slide rail topology network are identified. By optimizing the slide rail network, alternative paths can be added or the slide rail capacity can be expanded to alleviate these bottlenecks and ensure smooth operation of the trolley throughout the network. Simultaneously, redundant paths can be designed to ensure that if a slide rail fails or becomes congested, the trolley can continue to perform its task via an alternative path. This design helps improve the system's reliability and responsiveness.

[0018] By optimizing the first slide rail topology network based on multiple loading and unloading nodes to generate a second slide rail topology network, the flexibility and fault tolerance of the entire slide rail system can be enhanced. At the same time, the optimization of the second slide rail topology network enables the trolley to complete container loading and unloading tasks faster and more stably, thereby improving the efficiency and quality of the overall operation.

[0019] Step 4: Construct a path planning model based on the second sliding rail topology network.

[0020] Specifically, a path planning model is constructed based on the second sliding rail topology network. The path planning model not only considers the complex decision-making problem of multi-objective optimization (such as path length, energy consumption, safety, etc.), but also has the ability to adaptively adjust and plan in real time, thereby greatly improving the efficiency and reliability of the container loading and unloading process. The path planning model includes a strategy function, which is used to make multi-objective optimization decisions.

[0021] Step 5: The slide rail system receives real-time loading and unloading task information, slide rail real-time monitoring information, and environmental real-time monitoring information. It inputs the real-time loading and unloading task information, the slide rail real-time monitoring information, and the environmental real-time monitoring information into the path planning model for path planning and outputs the loading and unloading planned path.

[0022] Specifically, the slide rail system receives real-time loading and unloading task information, real-time slide rail monitoring information, and real-time environmental monitoring information through a sensor network and data interface. The real-time loading and unloading task information includes detailed information on currently pending container loading and unloading tasks, such as container location, target loading and unloading location, task priority, and urgency. This information guides the trolley in selecting priority tasks and optimizing routes. The real-time slide rail monitoring information includes the trolley's current position, speed, load status, and the status of slide rail nodes (such as whether they are occupied, node congestion, and the trolley's movement on the slide rail). This helps the system dynamically adjust routes and avoid conflicts between trolleys and path congestion. The real-time environmental monitoring information includes weather conditions (such as temperature, humidity, and wind speed), road conditions (such as the condition of the slide rail surface), and detection information of surrounding obstacles. This information is crucial for ensuring the safety of trolley operation and the reliability of route planning.

[0023] Next, the real-time loading and unloading task information, the real-time monitoring information of the sliding rail, and the real-time environmental monitoring information are input into the path planning model for path planning. Based on the received real-time data, the path planning model uses a strategy function to evaluate the current state and calculates the optimal action path that the trolley should take in the current state, outputting the loading and unloading planned path. This path guides the trolley from the current state to the target node to complete the loading and unloading task. At the same time, the system sends this path to the trolley control unit, and the trolley will move according to the planned path to ensure that the task is completed efficiently and safely.

[0024] This invention provides an intelligent path planning method for container loading and unloading, which addresses the technical problems arising from the complex and ever-changing container loading and unloading environment. Traditional path planning methods struggle to balance multiple optimization objectives, leading to an inability to quickly respond to dynamic changes and accurately set appropriate loading and unloading paths. This results in path conflicts, resource waste, and negatively impacts the efficiency and quality of container loading and unloading. The method constructs a first sliding rail topology network by distributing multiple sliding rail nodes within the container loading and unloading area and connecting these nodes to form a path graph. Next, multiple loading and unloading nodes within the container loading and unloading area are identified, and the first sliding rail topology network is optimized using these nodes to construct a second sliding rail topology network. A path planning model is then built based on this second topology network. Finally, the sliding rail system receives real-time loading and unloading task information, real-time sliding rail monitoring information, and real-time environmental monitoring information, inputting these into the path planning model for path planning and outputting the planned loading and unloading path. This method can quickly adapt to the dynamic environment during container loading and unloading, efficiently handle complex multi-objective optimization decisions, thereby improving the scientific accuracy and effectiveness of container loading and unloading path setting, and significantly enhancing the efficiency, quality, and safety of container loading and unloading.

[0025] Furthermore, after constructing the first sliding rail topology network, as shown in the attached diagram... Figure 2 As shown, this application also includes: The first slide rail topology network is layered to determine a multi-layer slide rail topology network; spatial intersection nodes are identified according to the multi-layer slide rail topology network, and identified nodes are output; collision risk analysis is performed on the identified nodes, and collision risk indicators of each identified node are output; identified nodes with collision risk indicators greater than preset collision risk are selected from each identified node to be optimized, the identified nodes to be optimized are optimized, and the first slide rail topology network is updated.

[0026] Specifically, after constructing the first slide rail topology network, the first slide rail topology network is layered. The purpose of layering is to divide the slide rail network into multiple layers according to different operational requirements and the characteristics of slide rail nodes. Each layer corresponds to a specific operational scenario or functional requirement, including functional layering, priority layering, geographical sealing, etc., to determine a multi-layer slide rail topology network. Each layer has an independent topology structure and path planning logic, but the layers can cooperate with each other.

[0027] Next, spatial intersection nodes are identified based on the multi-layer slide rail topology network. Spatial intersection nodes refer to key points where slide rail paths intersect between multiple slide rail layers or within the same layer. Sliders may simultaneously pass through or stop at these intersection nodes, leading to potential collision risks. For example, based on the path information in the multi-layer slide rail topology network, the system scans the paths of each slide rail layer and identifies the intersection points between paths. These intersection points are the intersection nodes. The identified intersection nodes are further classified to determine their slide rail layer and the slide rail paths they connect to, outputting the identified nodes. Then, collision risk analysis is performed on the identified nodes. For example, a risk assessment model can be constructed to analyze the potential collision risk of each identified node. The model can consider multiple factors, such as load conditions, to calculate the collision risk of each identified node; the collision risk index for each identified node is output, reflecting the probability and severity of slider collisions at different intersection nodes.

[0028] Then, nodes with collision risk indices exceeding a preset collision risk are selected from the various marker nodes to be optimized. The preset collision risk can be determined based on system safety requirements, historical collision event statistics, and specific conditions of trolley operation. Further optimization is then performed on these marker nodes. Appropriate optimization strategies are selected based on the specific circumstances of each node, including limiting trolley speed, increasing path redundancy, adjusting trolley scheduling, and optimizing node design. Finally, the first trolley topology network is updated based on the optimized node information. By selecting marker nodes with collision risk indices exceeding a preset threshold and performing targeted optimization on these nodes, the collision risk in the trolley system can be significantly reduced. The updated first trolley topology network will be safer and more robust, better supporting trolley scheduling and path planning during container loading and unloading, ensuring the efficient operation of the entire system and the safety of personnel and equipment.

[0029] Furthermore, the slide rail system is connected to an image acquisition device, which is housed within the container. This application also includes: The image acquisition device identifies the cargo in the container and obtains cargo granularity indicators, which are used to measure the geometric size of the cargo in the container. Based on the cargo granularity indicators, a collision risk analysis is performed on each identification node, and the collision risk indicators of each identification node are output.

[0030] Specifically, the sliding rail system is connected to an image acquisition device, which is installed inside the container. The container is then labeled with cargo data obtained from the image acquisition device, and cargo granularity indicators are acquired. These indicators, reflecting the geometric dimensions of the cargo (including length, width, height, and total volume), are used to measure the potential impact of cargo on the sliding rail nodes and path during transport. Next, a collision risk analysis is performed on each labeled node based on these cargo granularity indicators. This involves matching the cargo granularity indicators with the labeled node information of the sliding rail network to identify high-risk nodes that may be affected by the cargo's geometric size. For example, if a node has limited space and large cargo dimensions, the collision risk of that node may increase, and collision risk indicators for each labeled node are output.

[0031] Furthermore, the slide rail system includes a sliding carrier for transporting containers, and this application also includes: A first task state is set, which is a sliding carrier state with cargo loaded; a second task state is set, which is a sliding carrier state without cargo loaded; based on the first task state, the second sliding rail topology network is path-identified, and unidirectional sliding rail paths and bidirectional sliding rail paths with a first identification group are determined; based on the second task state, the second sliding rail topology network is path-identified, and unidirectional sliding rail paths and bidirectional sliding rail paths with a second identification group are determined, wherein the identification methods of the first identification group and the second identification group are different.

[0032] Specifically, the sliding rail system includes a sliding carrier used for transporting containers. The sliding carrier is a device within the sliding rail system used for transporting containers; it moves along a predetermined path in the sliding rail network to complete the loading and transportation tasks of the containers. Next, a first task state and a second task state are set. The first task state is the sliding carrier state with cargo loaded, and the second task state is the sliding carrier state without cargo loaded. Further, based on the first task state, the second sliding rail topology network is path-identified, determining unidirectional and bidirectional sliding rail paths with a first identification group; based on the second task state, the second sliding rail topology network is path-identified, determining unidirectional and bidirectional sliding rail paths with a second identification group. The identification methods for the first and second identification groups are different. The path identification method can use color coding, symbol marking, or other visual identification methods, such as red marking indicating a unidirectional path suitable for loading cargo, and blue marking indicating a unidirectional path suitable for unloaded cargo. Through differentiated identification, sliding rail paths suitable for different task states can be clearly distinguished, avoiding confusion during the selection and scheduling of sliding rail paths.

[0033] Furthermore, this application also includes: A unidirectional slide rail path and a bidirectional slide rail path are determined, wherein the unidirectional slide rail path supports a path in one of the forward or return directions, and the bidirectional slide rail path supports a path in both the forward and return directions; a first planning constraint is generated based on the unidirectional slide rail path with a first identifier group and the bidirectional slide rail path; a second planning constraint is generated based on the unidirectional slide rail path with a second identifier group; the path planning model optimizes the loading and unloading planning path based on the first planning constraint or the second planning constraint.

[0034] Specifically, in a sliding rail system, determining unidirectional and bidirectional sliding rail paths and generating corresponding planning constraints ensures the efficient and safe operation of the sliding vehicle under different task conditions. First, the unidirectional and bidirectional sliding rail paths are determined. The unidirectional sliding rail path supports movement in either the forward or return direction; that is, it only supports the sliding vehicle's movement in one direction. In a sliding rail system, a unidirectional sliding rail path helps simplify traffic flow management and reduces the risk of collisions at intersections. The bidirectional sliding rail path supports movement in both the forward and return directions; that is, it allows the sliding vehicle to operate in both directions. This type of path is suitable for transportation tasks requiring frequent round trips, improving the scheduling flexibility of the sliding vehicles and the system's operational efficiency.

[0035] Next, based on the unidirectional and bidirectional slide rail paths with the first set of identifiers, first planning constraints are generated. For example, in the first task state, the trolley can only travel along a unidirectional path in a specific direction (e.g., only forward and not backward, or only backward and not forward) to ensure transportation stability and avoid conflicts. The trolley is allowed to run in both directions, but its speed, load capacity, etc., may be limited to prevent collisions and congestion. Based on the unidirectional and bidirectional slide rail paths with the second set of identifiers, second planning constraints are generated. For example, when the trolley is not loaded, it can travel along more flexible unidirectional paths, which may allow higher speeds or allow the trolley to reverse direction when necessary to improve efficiency. Finally, the path planning model optimizes the loading and unloading planned path based on the first or second planning constraints.

[0036] Furthermore, a pressure sensor is disposed within the sliding carrier, and this application also includes:

[0037] Based on the pressure sensor, real-time pressure data of each sliding carrier is acquired; based on the real-time pressure data, the task status of each sliding carrier is determined. If the sliding carrier is in a first task status, the path planning model optimizes the loading and unloading planned path based on the first planning constraint; if the sliding carrier is in a second task status, the path planning model optimizes the loading and unloading planned path based on the second planning constraint.

[0038] Specifically, a pressure sensor is installed inside the sliding carrier to monitor the pressure on the carrier in real time. This data reflects the load status of the sliding carrier, thus determining whether it is currently loaded with goods. Next, based on the pressure sensor, real-time pressure data of each sliding carrier is acquired. Further, the task status of each sliding carrier is determined based on this real-time pressure data. If the pressure detected by the pressure sensor exceeds a certain threshold, the system determines that the sliding carrier is in a loaded state (first task state). If the pressure detected by the pressure sensor is below the set threshold, the system determines that the sliding carrier is in an unloaded state (second task state). If the sliding carrier is in the first task state, the path planning model optimizes the loading and unloading planned path based on the first planning constraints. For example, when the system determines that the sliding carrier is in the first task state (i.e., loaded with goods), the path planning model will prioritize more stable and wider paths and may also limit the speed of the trolleys to ensure the safety of goods transportation. If the sliding carrier is in the second task state, the path planning model optimizes the loading and unloading planning path based on the second planning constraints. For example, when the system determines that the sliding carrier is in the second task state (i.e., without cargo), the model can choose a more flexible and faster path, allowing the trolley to run at a higher speed on the path to improve task execution efficiency.

[0039] By installing pressure sensors within the sliding carrier, the slide rail system can acquire real-time pressure data and determine its task status. The path planning model then optimizes the path based on the carrier's task status, applying different planning constraints. This intelligent path optimization process not only improves the carrier's operating efficiency but also effectively ensures the safety of cargo transportation, guaranteeing efficient and reliable system operation under various task conditions.

[0040] Furthermore, this application also includes inputting the real-time loading and unloading task information, the real-time rail monitoring information, and the real-time environmental monitoring information into the path planning model for path planning. The real-time loading and unloading task information includes container location, target loading and unloading location, and cargo priority task order. The real-time rail monitoring information includes sliding carrier location, rail status, and node load. The real-time environmental monitoring information includes obstacle detection and temperature and humidity. The real-time loading and unloading task information, the real-time rail monitoring information, and the real-time environmental monitoring information are input into the path planning model, which includes a strategy function. The strategy function is introduced, and the strategy optimization result is output by maximizing the cumulative reward to obtain the loading and unloading planned path. The expression for strategy optimization includes: ;in, It is the policy parameter vector at the k-th iteration. It is the policy parameter vector at the (k+1)th iteration. The learning rate determines the magnitude of each parameter update. Calculate the current policy parameters The gradient of the expected return E, where E is the expected return. Given state s, the current policy parameters Corresponding strategy The probability of choosing action 'a', the strategy It is a probability distribution function. Given state s, the old policy parameters Corresponding strategy The probability of choosing action a For strategy ratio, Let be the advantage function, representing the superiority or inferiority of taking action a in state s compared to the average policy.

[0041] Specifically, the real-time loading and unloading task information includes container location, target loading and unloading location, and cargo priority task sequence; the real-time rail monitoring information includes sliding carrier location, rail status, and node load status; the rail status includes whether the rail path is unobstructed, whether there are obstacles, and whether the rail nodes are occupied or blocked; the real-time environmental monitoring information includes obstacle detection and temperature and humidity; the obstacle detection information includes whether there are physical obstacles on the rail path or nodes, which may hinder the normal passage of the trolley. The system needs to monitor the location of these obstacles in real time and avoid selecting these obstructed paths in path planning to ensure the safe operation of the trolley; temperature and humidity are important environmental factors affecting the safety of the trolley and cargo.

[0042] Next, the real-time loading / unloading task information, the real-time monitoring information of the slide rail, and the real-time environmental monitoring information are input into the path planning model, which includes a strategy function. Further, path planning optimization is performed based on the strategy function, i.e., maximizing the cumulative reward by selecting a series of actions to maximize the cumulative reward of the sled from the starting point to the end point. The optimization process involves repeated updates of the strategy function. In each iteration, the system calculates the actual reward based on the path planning result output by the current strategy and adjusts the parameters of the strategy function to increase the probability of selecting a high-reward path in the future. After multiple iterations of optimization, the path planning model outputs the strategy optimization result, obtaining the loading / unloading planned path, i.e., the optimal path selection that the sled should take under the current task state and environmental conditions. By introducing a strategy function and maximizing the cumulative reward for strategy optimization, the system can generate the optimal loading / unloading planned path. This method ensures that the sled can complete tasks efficiently and safely in complex and changing environments, while improving the overall performance of the system.

[0043] In the policy optimization function It is the policy parameter vector at the k-th iteration. It is the policy parameter vector at the (k+1)th iteration. The learning rate determines the magnitude of each parameter update. Calculate the current policy parameters The gradient of the expected return E, where E is the expected return. Given state s, the current policy parameters Corresponding strategy The probability of choosing action 'a', the strategy It is a probability distribution function. Given state s, the old policy parameters Corresponding strategy The probability of choosing action a For strategy ratio, Let be the advantage function, representing the merits of taking action 'a' relative to the average strategy in state 's'. By constructing a strategy optimization function, the system can provide a strong basis for the target optimization decision of the trolleys in the rail network. Maximizing cumulative rewards, utilizing the advantage function, and adjusting the strategy ratio ensures that each trolley can complete the loading and unloading task along the optimal path under complex task and environmental conditions. This optimization decision-making mechanism not only improves operational efficiency but also enhances the system's flexibility and adaptability.

[0044] Furthermore, this application also includes: Define the interface data between the slide rail system and the system driving the path planning model, including the data exchange format, data storage mode, and communication protocol; input the real-time loading and unloading task information, the real-time monitoring information of the slide rail, and the real-time environmental monitoring information into the path planning model according to the interface data.

[0045] Specifically, firstly, the interface data between the slide rail system and the system driving the path planning model is defined, including the data exchange format, data storage mode, and communication protocol. The data exchange format refers to the format in which data is transmitted between the slide rail system and the path planning model. The data storage mode determines how the slide rail system and the path planning model store and manage data. The communication protocol is the set of rules for data transmission between the slide rail system and the path planning model, determining how data is transmitted between the systems. Finally, based on the interface data, the real-time loading and unloading task information, the real-time slide rail monitoring information, and the real-time environmental monitoring information are input into the path planning model.

[0046] By defining the interface data between the slide rail system and the path planning model, the system can ensure that real-time loading and unloading task information, real-time slide rail monitoring information, and real-time environmental monitoring information can be accurately input into the path planning model. This precise data interface design ensures that the path planning algorithm can receive the system's real-time status information in a timely and accurate manner, thereby making optimal decisions and improving the overall performance and reliability of the system.

[0047] In summary, the intelligent path planning method for container loading and unloading provided in this application has the following technical effects: By distributing multiple sliding rail nodes in the container loading and unloading area, and connecting these nodes with sliding rails to form a path graph, a first sliding rail topology network is constructed. Next, multiple loading and unloading nodes in the container loading and unloading area are determined, and the first sliding rail topology network is optimized using these nodes to construct a second sliding rail topology network. Then, a path planning model is built based on the second sliding rail topology network. Finally, real-time loading and unloading task information, real-time sliding rail monitoring information, and real-time environmental monitoring information are received by the sliding rail system and input into the path planning model for path planning, outputting the planned loading and unloading path. This method can quickly adapt to the dynamically changing environment during container loading and unloading, efficiently handle complex decision-making problems involving multi-objective optimization, thereby improving the scientific nature and accuracy of container loading and unloading path setting, and effectively enhancing the efficiency, quality, and safety of container loading and unloading.

[0048] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0049] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for intelligent path planning in the container loading and unloading process, characterized in that, The methods include: Multiple slide rail nodes are distributed in the container loading and unloading area. The multiple slide rail nodes are connected by slide rails to form a path map and construct a first slide rail topology network. The first slide rail topology network is controlled by a slide rail system. Multiple loading and unloading nodes are identified in the container loading and unloading area, including loading and unloading platform nodes, container yard nodes, and forklift loading and unloading nodes; The first slide rail topology network is optimized using the multiple loading and unloading nodes to construct a second slide rail topology network; Based on the second sliding rail topology network, a path planning model is constructed; The slide rail system receives real-time loading and unloading task information, real-time slide rail monitoring information, and real-time environmental monitoring information. It inputs the real-time loading and unloading task information, the real-time slide rail monitoring information, and the real-time environmental monitoring information into the path planning model for path planning and outputs the loading and unloading planned path.

2. The intelligent path planning method for container loading and unloading process as described in claim 1, characterized in that, The slide rail system includes a sliding carrier for transporting containers, and the method further includes: Set the first task state, which is the sliding carrier state of cargo loading; Set a second task state, which is a sliding carrier state where the cargo is not loaded; Based on the first task status, the second slide rail topology network is path-identified to determine the unidirectional slide rail path and the bidirectional slide rail path with the first identification group; Based on the second task status, the second slide rail topology network is path-identified, and unidirectional slide rail paths and bidirectional slide rail paths with the second identification group are determined, wherein the identification methods of the first identification group and the second identification group are different.

3. The intelligent path planning method for container loading and unloading process as described in claim 2, characterized in that, Determine unidirectional and bidirectional slide rail paths, wherein the unidirectional slide rail path supports a path in one of the forward or return directions, and the bidirectional slide rail path supports a path in both the forward and return directions. Generate the first planning constraints based on the unidirectional and bidirectional sliding rail paths with the first identifier group; Generate the second planning constraints based on the unidirectional and bidirectional sliding rail paths with the second identifier group; The path planning model optimizes the loading and unloading planning path based on the first planning constraint or the second planning constraint.

4. The intelligent path planning method for container loading and unloading process as described in claim 3, characterized in that, The method further includes: A pressure sensor is installed inside the sliding carrier. Based on the pressure sensor, real-time pressure data of each sliding carrier is obtained; The task status of each sliding carrier is determined based on the real-time pressure data. If the sliding carrier is in the first task status, the path planning model optimizes the loading and unloading planning path based on the first planning constraint. If the sliding carrier is in the second task state, the path planning model optimizes the loading and unloading planning path based on the second planning constraints.

5. The intelligent path planning method for container loading and unloading process as described in claim 1, characterized in that, After constructing the first sliding rail topology network, the method also includes: The first slide rail topology network is layered to determine a multi-layer slide rail topology network; Based on the multi-layer sliding rail topology network, spatial intersection nodes are identified, and the identified nodes are output. Collision risk analysis is performed on the identified nodes, and the collision risk index of each identified node is output. Select the identification nodes that have a collision risk index greater than the preset collision risk from each identification node, optimize the identification nodes to be optimized, and update the first slide rail topology network.

6. The intelligent path planning method for container loading and unloading process as described in claim 5, characterized in that, The slide rail system is connected to an image acquisition device, which is housed within the container. The method includes: The image acquisition device identifies the cargo in the container and obtains cargo granularity indicators, which are used to measure the geometric size of the cargo in the container. Based on the cargo granularity index, a collision risk analysis is performed on each identification node, and the collision risk index of each identification node is output.

7. The intelligent path planning method for container loading and unloading process as described in claim 1, characterized in that, The real-time loading and unloading task information, the real-time rail monitoring information, and the real-time environmental monitoring information are input into the path planning model for path planning. include: The real-time loading and unloading task information includes container location, target loading and unloading location, and cargo priority task order; the real-time rail monitoring information includes sliding carrier location, rail status, and node load; and the real-time environmental monitoring information includes obstacle detection and temperature and humidity. The real-time loading and unloading task information, the real-time monitoring information of the slide rail, and the real-time environmental monitoring information are input into the path planning model, which includes a strategy function. Introducing the aforementioned policy function, the policy optimization result is output by maximizing the cumulative reward, resulting in the loading and unloading planning path. The policy optimization expression includes: ； in, It is the policy parameter vector at the k-th iteration. It is the policy parameter vector at the (k+1)th iteration. The learning rate determines the magnitude of each parameter update. Calculate the current policy parameters The gradient of the expected return E, where E is the expected return. Given state s, the current policy parameters Corresponding strategy The probability of choosing action 'a', the strategy It is a probability distribution function. Given state s, the old policy parameters Corresponding strategy The probability of choosing action a For strategy ratio, Let be the advantage function, representing the superiority or inferiority of taking action a in state s compared to the average policy.

8. The intelligent path planning method for container loading and unloading process as described in claim 1, characterized in that, Define the interface data between the slide rail system and the system driving the path planning model, including the data exchange format, data storage mode, and communication protocol; The real-time loading and unloading task information, the real-time rail monitoring information, and the real-time environmental monitoring information are input into the path planning model based on the interface data.

Images