PC-based intelligent transfer robot AGV path planning and control system

Abstract

The invention discloses an intelligent transfer robot AGV path planning and control system based on a PC, and relates to the technical field of robot control, the system comprises a data monitoring module, an AI dynamic path planning model module, a task planning module, an AGV path traffic monitoring module and an AGV intelligent obstacle avoidance module; environment and state data of the AGV are obtained in real time through various sensors and a 5G network, a dynamic three-dimensional space map is constructed, and a data set is generated; performing dynamic planning and optimization on the path based on a convolutional neural network; task allocation and path planning are carried out by using an ant colony algorithm, and task execution capability is evaluated; the traffic condition is monitored in real time, the congestion risk is evaluated, and the path is adjusted; and an avoidance coefficient is calculated during congestion, so that the AGV can pass safely, and the path is re-planned dynamically. According to the system, the task scheduling and path planning of the AGV are optimized, and the task execution efficiency and safety are improved.

Description

Technical Field

[0001] This invention relates to the field of robot control technology, specifically to a PC-based intelligent handling robot AGV path planning and control system. Background Technology

[0002] With the rapid development of intelligent manufacturing and logistics automation, Automated Guided Vehicles (AGVs) are increasingly widely used in warehousing, manufacturing, and retail. However, existing AGV path planning and control systems still suffer from many problems, such as low path planning efficiency, uneven task allocation, insufficient intelligence, and poor scalability.

[0003] Traditional AGVs often employ fixed paths or simple obstacle avoidance algorithms, making them ill-suited for complex environments such as multiple AGVs operating simultaneously or dynamic obstacles, resulting in inefficient path planning. Existing systems typically rely on static task allocation strategies, lacking the ability to dynamically adjust to real-time task load and AGV status, leading to some AGVs being overloaded while others remain idle. Existing systems lack real-time global environment awareness and prediction capabilities, making it difficult to reduce congestion and waiting time through path optimization. Furthermore, existing systems struggle to support collaborative operations in large-scale AGV clusters and cannot flexibly adapt to changing business needs. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a PC-based intelligent handling robot AGV path planning and control system to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solution: a PC-based intelligent handling robot AGV path planning and control system, including a data monitoring module, an AI dynamic path planning model module, a task planning module, an AGV path traffic monitoring module, and an AGV intelligent obstacle avoidance module; The data monitoring module is used to acquire real-time three-dimensional spatial information, location information and traffic data of the environment in which the AGV is located by using a lidar, ultrasonic sensor, infrared sensor, camera perspective recognition system and positioning and navigation system equipped with an AGV, and to construct a dynamic three-dimensional spatial map of the environment; to monitor the operating status of the AGV in real time, to upload data in real time using a 5G network, and to construct a first dataset and a second dataset; The AI ​​dynamic path planning model module is used to build an initial model using a convolutional neural network, and to train and test it using historical data to form an AI dynamic path planning model. By using the intermediate layer feature vectors in the first and second datasets, feature information is identified, and the AI ​​dynamic path planning model is continuously optimized to achieve real-time dynamic planning of AGV paths. The task planning module is used to dynamically allocate tasks using the ant colony algorithm, and combined with A The algorithm performs optimal path planning and calculates the task acceptance coefficient (CGL) to evaluate the AGV task execution capability, thereby realizing intelligent scheduling and risk warning of AGV tasks. The AGV path traffic monitoring module is used to monitor the traffic conditions in the AGV driving area in real time, calculate the traffic congestion coefficient JTY, and evaluate the traffic status by comparing it with the second threshold Q2, so as to realize early warning of AGV congestion risk and dynamic adjustment of the path. The AGV intelligent obstacle avoidance module is used to determine the priority right of passage of AGVs by calculating the AGV avoidance coefficient BRX when there is traffic congestion in the AGV's driving area, and to perform dynamic obstacle avoidance and path replanning for AGVs without priority.

[0006] Preferably, the data monitoring module includes a deployment unit and a data acquisition unit; The deployment unit is used to acquire real-time three-dimensional spatial information, location information, and traffic data of the AGV's environment by equipping the AGV with lidar, ultrasonic sensors, infrared sensors, camera perspective recognition system, and positioning and navigation system, and to construct a dynamic three-dimensional spatial map of the environment; and to upload the AGV's location information and traffic data to the PC dispatch center in real time based on 5G wireless network communication; and to establish a first dataset based on the three-dimensional spatial information, location information, and traffic data. The acquisition unit is used to install encoders to measure the wheel rotation of the AGV and acquire the speed of the AGV; to acquire the load data of the AGV by installing pressure sensors; to acquire the battery data and energy consumption data of the AGV by monitoring the battery status of the AGV; and to establish a second dataset.

[0007] Preferably, the AI ​​dynamic path planning model module is used to construct an initial convolutional neural network model using a convolutional neural network, train and test the initial convolutional neural network model with historical data, and use the trained initial convolutional neural network model as the AI ​​dynamic path planning model. At the same time, the intermediate layer output of the first dataset and the second dataset is used as a feature vector to identify feature information. The AI ​​dynamic path planning model is trained and tested using the acquired feature information, and the trained AI dynamic path planning model module is run as data to perform real-time dynamic path planning for the AGV.

[0008] Preferably, the task planning module includes a first calculation unit, a task issuing unit, a second calculation unit, a third calculation unit, and a first analysis unit; The first computing unit is used for a dynamic task allocation mechanism based on the ant colony algorithm. Combining data from the first dataset and the second dataset, and after dimensionless processing, it calculates and obtains the task level coefficient RWX of the intelligent handling robot, as shown in the following formula: ; In the formula, qw represents the task weight score, zp represents the probability of historical task completion, D represents the path congestion coefficient, L represents the AGV load, E represents the task energy consumption, V represents the AGV speed, and T represents the remaining task time. and Indicates the weighting coefficient; ; In the formula, NA represents the number of AGVs running on the path, Lpath represents the path length, V represents the speed of the AGV, and Wpath represents the bandwidth. The task issuing unit is used to receive task signals from the PC, task pick-up signals from the AGV, unpicked task signals from the AGV, and task completion signals from the AGV. It then calculates and obtains the task level coefficient RWX of the intelligent handling robot, and sends task assignment signals to idle AGVs in descending order of RWX value.

[0009] Preferably, the second calculation unit is used to, when the AGV receives the task deployment signal, combine it with A The algorithm, after dimensionless processing, calculates and obtains the optimal path from the current position to the task's end point. The formula is as follows: ; In the formula, p1 represents the set of path nodes from the starting point to the task point, and n represents the number of path nodes. This represents the distance between any two nodes on the path; The third calculation unit is used to calculate the optimal path from the current position to the task end point. Combining the data from the second dataset, after dimensionless processing, the task acceptance coefficient CGL is calculated and obtained, as shown in the following formula: ; In the formula, Let L represent the task level coefficient of the j-th task, L represent the load of the AGV, and ER represent the power status of the AGV.

[0010] Preferably, the first analysis unit is used to pre-set a first threshold Q1 and compare the task acceptance coefficient CGL with the first threshold Q1 to obtain a first evaluation result, including: When the task acceptance coefficient CGL < the first threshold Q1, it indicates that the AGV is at risk of not being able to complete the j-th task, triggering the first warning instruction and generating the first strategy: the AGV abandons receiving the task and sends an unaccepted task signal to the task issuing unit. When the task acceptance coefficient CGL ≥ the first threshold Q1, it means that the AGV has no risk of not being able to complete the j-th task. The AGV automatically picks up the task and starts execution, and sends a task pick-up signal to the task issuing unit. After completing the task, it sends a task completion signal to the task issuing unit.

[0011] Preferably, the AGV path traffic monitoring module includes a fourth calculation unit and a second analysis unit; The fourth calculation unit is used to monitor traffic conditions in the AGV's operating area. Combining data from the first and second datasets, and after dimensionless processing, it calculates and obtains the traffic congestion coefficient JTY, as shown in the following formula: ; In the formula, NTA represents the number of AGVs in the operating area. Indicates the maximum capacity of the driving area. This indicates the path length occupied by the h-th AGV. This indicates the maximum length of path that can be occupied within the driving area. This represents the average speed of the AGV, and 'a' represents the adjustment coefficient. The second analysis unit is used to pre-set a second threshold Q2 and compare the traffic congestion coefficient JTY with the second threshold Q2 to obtain a second evaluation result, including: When the traffic congestion coefficient JTY < the second threshold Q2, it indicates that the traffic in the AGV's operating area is smooth and there is no risk of congestion. Continuous monitoring is required. When the traffic congestion coefficient JTY is greater than or equal to the second threshold Q2, it indicates that the traffic in the AGV's operating area is busy and there is a risk of congestion. This triggers the second warning instruction and generates the second strategy: issuing avoidance instructions to some AGVs and readjusting their paths.

[0012] Preferably, the AGV intelligent obstacle avoidance module includes a fifth calculation unit, a third analysis unit, and a sixth calculation unit; The fifth calculation unit is used to monitor the task completion level of the AGV when the second early warning command is received, and calculates the AGV avoidance coefficient BRX after dimensionless processing, combining the task level coefficient with the coefficient. The formula is as follows: ; In the formula, JTY represents the traffic congestion coefficient. This represents the task level coefficient of the b-th AGV. This indicates the path length occupied by the b-th AGV. This indicates the completion status of the task for the b-th AGV.

[0013] Preferably, the third analysis unit is used to pre-set a third threshold Q3 and compare the AGV avoidance coefficient BRX with the third threshold Q3 to obtain a third evaluation result, including: When the AGV avoidance coefficient BRX ≤ the third threshold Q3, it means that the current AGV has priority passage and does not need to avoid, and continuous monitoring is required; When the AGV avoidance coefficient BRX is greater than the third threshold Q3, it means that the current AGV does not have priority passage and needs to avoid it, triggering the third warning instruction.

[0014] Preferably, the sixth calculation unit is used to regenerate a new path for AGVs with an AGV avoidance coefficient BRX > third threshold Q3 when a third warning instruction is received. The formula is as follows: ; In the formula, p2 represents the set of path nodes from the obstacle avoidance start point to the task end point, and m represents the number of path nodes. Indicates the cost of the new path. This represents the time cost of the i-th node. This represents the obstacle avoidance cost of the i-th node.

[0015] This invention provides a PC-based intelligent handling robot (AGV) path planning and control system. It has the following advantages: (1) This PC-based intelligent handling robot AGV path planning and control system utilizes convolutional neural network (CNN) and real-time dataset to perform dynamic path planning for the AGV, thereby achieving automatic path adjustment according to environmental changes. This AI-based dynamic path planning model can optimize the AGV's driving route in real time, reducing path planning time and congestion risks, and improving system operating efficiency, especially in complex environments, such as multiple AGVs operating in parallel or the presence of dynamic obstacles.

[0016] (2) The PC-based intelligent handling robot AGV path planning and control system combines ant colony algorithm and A* algorithm. The present invention realizes dynamic task allocation and path planning based on real-time data. It can not only evaluate the execution capability of the task according to the priority of the task, the load and status of the AGV, but also adjust the task allocation in real time to avoid AGV overload or idleness, effectively improving the efficiency of resource utilization and task completion rate.

[0017] (3) This PC-based intelligent handling robot AGV path planning and control system monitors the traffic conditions in the AGV's operating area in real time, calculates the traffic congestion coefficient, and evaluates it in conjunction with a set threshold. When a traffic congestion risk occurs, it can automatically trigger an early warning command and adjust the path to avoid potential congestion and reduce waiting time. This intelligent monitoring system enhances the adaptability and stability of the AGV system in large-scale environments.

[0018] (4) This PC-based intelligent transport robot AGV path planning and control system can monitor the AGV's task completion status and obstacle avoidance coefficient in real time through the AGV intelligent obstacle avoidance module. Through dynamic calculation and comparison, it can reasonably determine the AGV's priority right of passage. When traffic congestion occurs, the system will automatically generate an obstacle avoidance path, thereby ensuring that the AGV completes its task efficiently and safely with limited resources. This intelligent obstacle avoidance control system effectively avoids collisions and unnecessary delays, improving the system's safety and operating efficiency. Attached Figure Description

[0019] Figure 1 This is a block diagram flowchart of a PC-based intelligent handling robot AGV path planning and control system according to the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Example 1 Please see Figure 1 This invention provides a PC-based intelligent handling robot AGV path planning and control system, including a data monitoring module, an AI dynamic path planning model module, a task planning module, an AGV path traffic monitoring module, and an AGV intelligent obstacle avoidance module. The data monitoring module is used to acquire real-time three-dimensional spatial information, location information and traffic data of the environment in which the AGV is located by using a lidar, ultrasonic sensor, infrared sensor, camera perspective recognition system and positioning and navigation system equipped with an AGV, and to construct a dynamic three-dimensional spatial map of the environment; to monitor the operating status of the AGV in real time, to upload data in real time using a 5G network, and to construct a first dataset and a second dataset; The AI ​​dynamic path planning model module is used to build an initial model using a convolutional neural network, and to train and test it using historical data to form an AI dynamic path planning model. By using the intermediate layer feature vectors in the first and second datasets, feature information is identified, and the AI ​​dynamic path planning model is continuously optimized to achieve real-time dynamic planning of AGV paths. The task planning module is used to dynamically allocate tasks using the ant colony algorithm, and combined with A The algorithm performs optimal path planning and calculates the task acceptance coefficient (CGL) to evaluate the AGV task execution capability, thereby realizing intelligent scheduling and risk warning of AGV tasks. The AGV path traffic monitoring module is used to monitor the traffic conditions in the AGV driving area in real time, calculate the traffic congestion coefficient JTY, and evaluate the traffic status by comparing it with the second threshold Q2, so as to realize early warning of AGV congestion risk and dynamic adjustment of the path. The AGV intelligent obstacle avoidance module is used to determine the priority right of passage of AGVs by calculating the AGV avoidance coefficient BRX when there is traffic congestion in the AGV's driving area, and to perform dynamic obstacle avoidance and path replanning for AGVs without priority.

[0022] In this embodiment, by integrating multiple modules such as data monitoring, AI dynamic path planning, task planning, path traffic monitoring, and intelligent obstacle avoidance, comprehensive monitoring and intelligent adjustment of the AGV system are achieved. The data monitoring module acquires the 3D spatial and location information and traffic data of the AGV environment in real time, and quickly uploads and constructs a dataset in real time via the 5G network. The AI ​​dynamic path planning module continuously optimizes the path based on historical data to ensure real-time dynamic adjustment of the path. The task planning module uses ant colony optimization and AI... The algorithm intelligently allocates tasks and plans paths, while simultaneously evaluating the AGV's task execution capabilities to reduce task overload or idleness. The traffic monitoring module and intelligent obstacle avoidance module work together to monitor traffic conditions in real time and issue congestion warnings, ensuring that the AGV can safely and smoothly complete tasks in busy environments. Through these intelligent system controls, the operating efficiency and safety of AGVs can be significantly improved, the risks of congestion and collisions reduced, and efficient collaboration of large-scale AGV clusters in complex environments guaranteed.

[0023] Example 2, This embodiment is an explanation of Embodiment 1. Specifically, the data monitoring module includes a deployment unit and a data acquisition unit. The deployment unit is used to acquire real-time three-dimensional spatial information, location information, and traffic data of the AGV's environment by equipping the AGV with lidar, ultrasonic sensors, infrared sensors, camera perspective recognition system, and positioning and navigation system, and to construct a dynamic three-dimensional spatial map of the environment; and to upload the AGV's location information and traffic data to the PC dispatch center in real time based on 5G wireless network communication; and to establish a first dataset based on the three-dimensional spatial information, location information, and traffic data. The acquisition unit is used to install encoders to measure the wheel rotation of the AGV and acquire the speed of the AGV; to acquire the load data of the AGV by installing pressure sensors; to acquire the battery data and energy consumption data of the AGV by monitoring the battery status of the AGV; and to establish a second dataset.

[0024] In this embodiment, the deployment unit and the data acquisition unit work together to comprehensively and in real-time perceive the environment in which the AGV is located. The deployment unit uses lidar, ultrasonic sensors, infrared sensors, a camera perspective recognition system, and a positioning and navigation system to accurately acquire 3D spatial information, location information, and traffic data of the environment, constructing a dynamic 3D spatial map of the environment. This data is then uploaded to the PC dispatch center in real-time via a 5G wireless network, ensuring timely information transmission. The data acquisition unit focuses on real-time monitoring of the AGV's speed, load, battery status, and energy consumption, generating a second dataset. This comprehensive data acquisition and real-time uploading capability not only improves the AGV's environmental adaptability but also enhances its task execution efficiency.

[0025] Example 3 is an explanation of Example 2. Please refer to the example provided. Figure 1 Specifically, the AI ​​dynamic path planning model module is used to construct an initial convolutional neural network model using a convolutional neural network, train and test the initial convolutional neural network model with historical data, and use the trained initial convolutional neural network model as the AI ​​dynamic path planning model. At the same time, the intermediate layer output of the first dataset and the second dataset is used as a feature vector to identify feature information. The AI ​​dynamic path planning model is trained and tested using the acquired feature information, and the trained AI dynamic path planning model module is run as data to perform real-time dynamic path planning for the AGV.

[0026] In this embodiment, the present invention utilizes an AI dynamic path planning model module to intelligently optimize path planning using a convolutional neural network (CNN). This module constructs an initial CNN model by training on historical data and continuously optimizes the AI ​​model by combining intermediate layer feature vectors from the first and second datasets. This model can identify environmental feature information in real time and dynamically adjust the AGV's path planning, thereby achieving efficient and flexible path selection. Through this AI-based dynamic learning and optimization, the AGV can autonomously adapt to changing road conditions and obstacles in complex environments, achieving accurate and efficient navigation, significantly improving the system's operating efficiency and the intelligence level of path planning.

[0027] Example 4 is an explanation of Example 3. Please refer to the example provided. Figure 1 Specifically, the task planning module includes a first calculation unit, a task issuing unit, a second calculation unit, a third calculation unit, and a first analysis unit; The first computing unit is used for a dynamic task allocation mechanism based on the ant colony algorithm. Combining data from the first dataset and the second dataset, and after dimensionless processing, it calculates and obtains the task level coefficient RWX of the intelligent handling robot, as shown in the following formula: ; In the formula, qw represents the task weight score, as shown in the table below; zp represents the probability of historical task completion; D represents the path congestion coefficient; L represents the AGV load; E represents the task energy consumption; V represents the AGV speed; and T represents the remaining task time. and Indicates the weighting coefficient; ; ; In the formula, NA represents the number of AGVs running on the path, Lpath represents the path length, V represents the speed of the AGV, and Wpath represents the bandwidth. The task issuing unit is used to receive task signals from the PC, task pick-up signals from the AGV, unpicked task signals from the AGV, and task completion signals from the AGV. It then calculates and obtains the task level coefficient RWX of the intelligent handling robot, and sends task assignment signals to idle AGVs in descending order of RWX value.

[0028] In this embodiment, a dynamic task allocation mechanism based on the ant colony algorithm achieves efficient task allocation and scheduling. The first computing unit combines the first and second datasets and calculates the task priority coefficient RWX after dimensionless processing to intelligently evaluate tasks. Based on factors such as task priority, path congestion, and AGV load, the system can dynamically adjust task allocation, ensuring that tasks are assigned to idle AGVs in descending order of priority. This intelligent scheduling mechanism optimizes resource utilization, avoids overloading or idleness of some AGVs, improves the overall task execution efficiency of the system, and ensures that urgent tasks are prioritized, effectively reducing task delays and resource waste.

[0029] Example 5 is an explanation of Example 4. Please refer to the example provided. Figure 1 Specifically, the second calculation unit is used to, when the AGV receives the task deployment signal, combine it with A The algorithm, after dimensionless processing, calculates and obtains the optimal path from the current position to the task's end point. The formula is as follows: ; In the formula, p1 represents the set of path nodes from the starting point to the task point, and n represents the number of path nodes. This represents the distance between any two nodes on the path; The third calculation unit is used to calculate the optimal path from the current position to the task end point. Combining the data from the second dataset, after dimensionless processing, the task acceptance coefficient CGL is calculated and obtained, as shown in the following formula: ; In the formula, Let L represent the task level coefficient of the j-th task, L represent the load of the AGV, and ER represent the power status of the AGV.

[0030] In this embodiment, the second computing unit and the third computing unit combine A The algorithm and dimensionless processing technology enable optimal path planning for the AGV from its current position to the task's endpoint, and accurately assess its task-taking capacity. The second calculation unit calculates the optimal path based on the task assignment signal, ensuring the AGV completes the task in the shortest time. The third calculation unit considers the AGV's load and power status to calculate the task-taking coefficient (CGL), thereby evaluating whether the AGV can successfully complete the task. By dynamically adjusting task allocation and path planning, the system not only improves the task completion rate but also optimizes the AGV's load and energy usage, ensuring the AGV's stability and efficiency during operation.

[0031] Example 6 is an explanation of Example 5. Please refer to the example provided. Figure 1 Specifically, the first analysis unit is used to pre-set a first threshold Q1, and compare the task acceptance coefficient CGL with the first threshold Q1 to obtain a first evaluation result, including: When the task acceptance coefficient CGL < the first threshold Q1, it indicates that the AGV is at risk of not being able to complete the j-th task, triggering the first warning instruction and generating the first strategy: the AGV abandons receiving the task and sends an unaccepted task signal to the task issuing unit. When the task acceptance coefficient CGL ≥ the first threshold Q1, it means that the AGV has no risk of not being able to complete the j-th task. The AGV automatically picks up the task and starts execution, and sends a task pick-up signal to the task issuing unit. After completing the task, it sends a task completion signal to the task issuing unit.

[0032] In this embodiment, the task acceptance coefficient CGL is compared and analyzed with a preset first threshold Q1 to evaluate the AGV's task execution capability in real time. When the task acceptance coefficient is lower than the first threshold, the system triggers an early warning mechanism, automatically abandoning task acceptance to prevent the AGV from failing to complete tasks due to overload or insufficient power, thereby reducing system failures and resource waste. When the task acceptance coefficient is higher than the threshold, the AGV automatically accepts tasks and begins execution, ensuring timely task completion. This intelligent decision-making mechanism optimizes the task allocation process, improves the stability and reliability of task execution, and ensures that the AGV operates efficiently without overloading.

[0033] In this embodiment, by comparing the task capacity coefficient CGL with the preset threshold Q1 in real time, the system ensures that the AGV will not operate under overload conditions when performing tasks. When the AGV's task capacity is insufficient, the system will automatically trigger an early warning and abandon task assignment, thereby avoiding task failure due to insufficient power or excessive load, reducing the risk of failure, and maximizing the system's operating efficiency and reliability. This mechanism effectively optimizes task allocation, avoids resource waste, and improves the AGV's adaptability in complex environments and its long-term operational stability.

[0034] Example 7 is an explanation of Example 6. Please refer to the example provided. Figure 1 Specifically, the AGV path traffic monitoring module includes a fourth calculation unit and a second analysis unit; The fourth calculation unit is used to monitor traffic conditions in the AGV's operating area. Combining data from the first and second datasets, and after dimensionless processing, it calculates and obtains the traffic congestion coefficient JTY, as shown in the following formula: ; In the formula, NTA represents the number of AGVs in the operating area. Indicates the maximum capacity of the driving area. This indicates the path length occupied by the h-th AGV. This indicates the maximum length of path that can be occupied within the driving area. This represents the average speed of the AGV, and 'a' represents an adjustment coefficient obtained experimentally to control the impact of average speed on traffic congestion. ; The second analysis unit is used to pre-set a second threshold Q2 and compare the traffic congestion coefficient JTY with the second threshold Q2 to obtain a second evaluation result, including: When the traffic congestion coefficient JTY < the second threshold Q2, it indicates that the traffic in the AGV's operating area is smooth and there is no risk of congestion. Continuous monitoring is required. When the traffic congestion coefficient JTY is greater than or equal to the second threshold Q2, it indicates that the traffic in the AGV's operating area is busy and there is a risk of congestion. This triggers the second warning instruction and generates the second strategy: issuing avoidance instructions to some AGVs and readjusting their paths.

[0035] In this embodiment, the traffic congestion coefficient JTY is dynamically calculated and compared with a preset second threshold Q2 through real-time traffic monitoring and analysis of the AGV path traffic monitoring module. When the traffic congestion coefficient exceeds the threshold, the system triggers an early warning and adjusts the paths of some AGVs in a timely manner to avoid delays and resource waste caused by traffic congestion. This mechanism effectively reduces the waiting time of AGVs in complex traffic environments, improves the traffic efficiency of the driving area, ensures that multiple AGVs can complete tasks efficiently and safely by cooperating with each other, and reduces the risk of traffic conflicts during system operation.

[0036] Example 8 is an explanation of Example 7. Please refer to the example provided. Figure 1 Specifically, the AGV intelligent obstacle avoidance module includes a fifth calculation unit, a third analysis unit, and a sixth calculation unit; The fifth calculation unit is used to monitor the task completion level of the AGV when the second early warning command is received, and calculates the AGV avoidance coefficient BRX after dimensionless processing, combining the task level coefficient with the coefficient. The formula is as follows: ; In the formula, JTY represents the traffic congestion coefficient. This represents the task level coefficient of the b-th AGV. This indicates the path length occupied by the b-th AGV. This indicates the completion status of the task for the b-th AGV.

[0037] In this embodiment, upon receiving a traffic congestion warning, the AGV intelligent obstacle avoidance module intelligently assesses the AGV's obstacle avoidance coefficient based on the task level coefficient and task completion status, and adjusts the AGV's travel path in real time. This mechanism ensures that in traffic congestion, the system prioritizes AGVs with higher task importance or those nearing completion, avoiding task delays or resource waste due to obstacle avoidance. This intelligent obstacle avoidance and path replanning function improves the AGV's adaptability and flexibility, enabling it to more efficiently avoid congested areas in complex environments, ensuring timely task completion and optimizing the overall system's operational efficiency.

[0038] Example 9 is an explanation of Example 8. Please refer to the example provided. Figure 1 Specifically, the third analysis unit is used to pre-set a third threshold Q3 and compare the AGV avoidance coefficient BRX with the third threshold Q3 to obtain the third evaluation result, including: When the AGV avoidance coefficient BRX ≤ the third threshold Q3, it means that the current AGV has priority passage and does not need to avoid, and continuous monitoring is required; When the AGV avoidance coefficient BRX is greater than the third threshold Q3, it means that the current AGV does not have priority passage and needs to avoid it, triggering the third warning instruction.

[0039] In this embodiment, the system intelligently determines the passage priority of each AGV by comparing the AGV avoidance coefficient with a preset third threshold Q3 through a third analysis unit. When the AGV avoidance coefficient is lower than or equal to the third threshold, the AGV does not need to avoid the obstacle and continues normal driving; however, when the avoidance coefficient exceeds the third threshold, the system will promptly trigger an early warning and instruct the AGV to avoid the obstacle. This mechanism ensures the smooth operation of AGVs in complex traffic environments. Through real-time dynamic adjustment and priority right-of-way determination, it optimizes the cooperation and smoothness between AGVs, reduces unnecessary path conflicts and congestion, and improves the overall system efficiency and stability.

[0040] Example 10 is an explanation of Example 9. Please refer to the example provided. Figure 1 Specifically, the sixth calculation unit is used to regenerate a new path for AGVs with an AGV avoidance coefficient BRX > third threshold Q3 when a third warning instruction is received. The formula is as follows: ; In the formula, p2 represents the set of path nodes from the obstacle avoidance start point to the task end point, and m represents the number of path nodes. Indicates the cost of the new path. This represents the time cost of the i-th node. This represents the obstacle avoidance cost of the i-th node.

[0041] In this embodiment, when the AGV's obstacle avoidance coefficient exceeds the third threshold Q3, triggering the third warning command, the sixth calculation unit can replan the AGV's travel path in real time. By recalculating the time cost and obstacle avoidance cost of the obstacle avoidance path, the system generates a new optimal path, ensuring that the AGV can safely and quickly pass through congested or obstacle-heavy areas. This dynamic path replanning mechanism can effectively avoid collisions and traffic jams, improve the AGV's obstacle avoidance capability in complex environments, ensure that tasks are completed on time and efficiently, and enhance the flexibility and safety of the AGV system.

[0042] The threshold is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by those skilled in the art for each set of sample data; as long as it does not affect the ratio between the parameter and the quantized value, it is acceptable.

[0043] The above formulas are all derived from software simulation using a large amount of data and are selected to be close to the actual values. The coefficients in the formulas are set by those skilled in the art according to the actual situation. The above description is only a preferred embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any equivalent substitutions or changes made by those skilled in the art within the technical scope disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the protection scope of the present invention.

Claims

1. A PC-based intelligent handling robot AGV path planning and control system, characterized in that, It includes a data monitoring module, an AI dynamic path planning model module, a task planning module, an AGV path traffic monitoring module, and an AGV intelligent obstacle avoidance module; The data monitoring module is used to acquire real-time three-dimensional spatial information, location information and traffic data of the environment in which the AGV is located by using a lidar, ultrasonic sensor, infrared sensor, camera perspective recognition system and positioning and navigation system equipped with an AGV, and to construct a dynamic three-dimensional spatial map of the environment; to monitor the operating status of the AGV in real time, to upload data in real time using a 5G network, and to construct a first dataset and a second dataset; The AI ​​dynamic path planning model module is used to build an initial model using a convolutional neural network, and to train and test it using historical data to form an AI dynamic path planning model. By using the intermediate layer feature vectors in the first and second datasets, feature information is identified, and the AI ​​dynamic path planning model is continuously optimized to achieve real-time dynamic planning of AGV paths. The task planning module is used to dynamically allocate tasks using the ant colony algorithm, and combined with A The algorithm performs optimal path planning and calculates the task acceptance coefficient (CGL) to evaluate the AGV task execution capability, thereby realizing intelligent scheduling and risk warning of AGV tasks. The AGV path traffic monitoring module is used to monitor the traffic conditions in the AGV driving area in real time, calculate the traffic congestion coefficient JTY, and evaluate the traffic status by comparing it with the second threshold Q2, so as to realize early warning of AGV congestion risk and dynamic adjustment of the path. The AGV intelligent obstacle avoidance module is used to determine the priority right of passage of AGVs by calculating the AGV avoidance coefficient BRX when there is traffic congestion in the AGV's driving area, and to perform dynamic obstacle avoidance and path replanning for AGVs without priority.

2. The PC-based intelligent handling robot AGV path planning and control system according to claim 1, characterized in that, The data monitoring module includes a deployment unit and a data acquisition unit; The deployment unit is used to acquire real-time three-dimensional spatial information, location information, and traffic data of the AGV's environment by equipping the AGV with lidar, ultrasonic sensors, infrared sensors, camera perspective recognition system, and positioning and navigation system, and to construct a dynamic three-dimensional spatial map of the environment; and to upload the AGV's location information and traffic data to the PC dispatch center in real time based on 5G wireless network communication; and to establish a first dataset based on the three-dimensional spatial information, location information, and traffic data. The acquisition unit is used to install encoders to measure the wheel rotation of the AGV and acquire the speed of the AGV; to acquire the load data of the AGV by installing pressure sensors; to acquire the battery data and energy consumption data of the AGV by monitoring the battery status of the AGV; and to establish a second dataset.

3. The PC-based intelligent handling robot AGV path planning and control system according to claim 2, characterized in that, The AI ​​dynamic path planning model module is used to construct an initial convolutional neural network model using a convolutional neural network, train and test the initial convolutional neural network model with historical data, and use the trained initial convolutional neural network model as the AI ​​dynamic path planning model. At the same time, the intermediate layer output of the first dataset and the second dataset is used as a feature vector to identify feature information. The AI ​​dynamic path planning model is trained and tested using the acquired feature information, and the trained AI dynamic path planning model module is run as data to perform real-time dynamic path planning for AGV.

4. The PC-based intelligent handling robot AGV path planning and control system according to claim 3, characterized in that, The task planning module includes a first calculation unit, a task issuing unit, a second calculation unit, a third calculation unit, and a first analysis unit; The first computing unit is used for a dynamic task allocation mechanism based on the ant colony algorithm. Combining data from the first dataset and the second dataset, and after dimensionless processing, it calculates and obtains the task level coefficient RWX of the intelligent handling robot, as shown in the following formula: ； In the formula, qw represents the task weight score, zp represents the probability of historical task completion, D represents the path congestion coefficient, L represents the AGV load, E represents the task energy consumption, V represents the AGV speed, and T represents the remaining task time. and Indicates the weighting coefficient; ； In the formula, NA represents the number of AGVs running on the path, Lpath represents the path length, V represents the speed of the AGV, and Wpath represents the bandwidth. The task issuing unit is used to receive task signals from the PC, task pick-up signals from the AGV, unpicked task signals from the AGV, and task completion signals from the AGV. It then calculates and obtains the task level coefficient RWX of the intelligent handling robot, and sends task assignment signals to idle AGVs in descending order of RWX value.

5. The PC-based intelligent handling robot AGV path planning and control system according to claim 4, characterized in that, The second calculation unit is used to, when the AGV receives the task deployment signal, combine it with A The algorithm, after dimensionless processing, calculates and obtains the optimal path from the current position to the task's end point. The formula is as follows: ； In the formula, p1 represents the set of path nodes from the starting point to the task point, and n represents the number of path nodes. This represents the distance between any two nodes on the path; The third calculation unit is used to calculate the optimal path from the current position to the task end point. Combining the data from the second dataset, after dimensionless processing, the task acceptance coefficient CGL is calculated and obtained, as shown in the following formula: ； In the formula, Let L represent the task level coefficient of the j-th task, L represent the load of the AGV, and ER represent the power status of the AGV.

6. The PC-based intelligent handling robot AGV path planning and control system according to claim 5, characterized in that, The first analysis unit is used to pre-set a first threshold Q1, and compare the task acceptance coefficient CGL with the first threshold Q1 to obtain a first evaluation result, including: When the task acceptance coefficient CGL < the first threshold Q1, it indicates that the AGV is at risk of not being able to complete the j-th task, triggering the first warning instruction and generating the first strategy: the AGV abandons receiving the task and sends an unaccepted task signal to the task issuing unit. When the task acceptance coefficient CGL ≥ the first threshold Q1, it means that the AGV has no risk of not being able to complete the j-th task. The AGV automatically picks up the task and starts execution, and sends a task pick-up signal to the task issuing unit. After completing the task, it sends a task completion signal to the task issuing unit.

7. The PC-based intelligent handling robot AGV path planning and control system according to claim 6, characterized in that, The AGV path traffic monitoring module includes a fourth calculation unit and a second analysis unit; The fourth calculation unit is used to monitor traffic conditions in the AGV's operating area. Combining data from the first and second datasets, and after dimensionless processing, it calculates and obtains the traffic congestion coefficient JTY, as shown in the following formula: ； In the formula, NTA represents the number of AGVs in the operating area. Indicates the maximum capacity of the driving area. This indicates the path length occupied by the h-th AGV. This indicates the maximum length of path that can be occupied within the driving area. This represents the average speed of the AGV, where 'a' represents the adjustment coefficient. The second analysis unit is used to pre-set a second threshold Q2 and compare the traffic congestion coefficient JTY with the second threshold Q2 to obtain a second evaluation result, including: When the traffic congestion coefficient JTY < the second threshold Q2, it indicates that the traffic in the AGV's operating area is smooth and there is no risk of congestion. Continuous monitoring is required. When the traffic congestion coefficient JTY is greater than or equal to the second threshold Q2, it indicates that the traffic in the AGV's operating area is busy and there is a risk of congestion. This triggers the second warning instruction and generates the second strategy: issuing avoidance instructions to some AGVs and readjusting their paths.

8. The PC-based intelligent handling robot AGV path planning and control system according to claim 7, characterized in that, The AGV intelligent obstacle avoidance module includes a fifth calculation unit, a third analysis unit, and a sixth calculation unit; The fifth calculation unit is used to monitor the task completion level of the AGV when the second early warning command is received, and calculates the AGV avoidance coefficient BRX after dimensionless processing, combining the task level coefficient with the coefficient. The formula is as follows: ； In the formula, JTY represents the traffic congestion coefficient. This represents the task level coefficient of the b-th AGV. This indicates the path length occupied by the b-th AGV. This indicates the completion status of the task for the b-th AGV.

9. The PC-based intelligent handling robot AGV path planning and control system according to claim 8, characterized in that, The third analysis unit is used to pre-set a third threshold Q3, and compare the AGV avoidance coefficient BRX with the third threshold Q3 to obtain the third evaluation result, including: When the AGV avoidance coefficient BRX ≤ the third threshold Q3, it means that the current AGV has priority passage and does not need to avoid the obstacle; continuous monitoring is required. When the AGV avoidance coefficient BRX is greater than the third threshold Q3, it means that the current AGV does not have priority passage and needs to avoid it, triggering the third warning instruction.

10. A PC-based intelligent handling robot AGV path planning and control system according to claim 9, characterized in that, The sixth calculation unit is used to regenerate a new path for AGVs with an AGV avoidance coefficient greater than the third threshold Q3 when a third warning instruction is received. The formula is as follows: ； In the formula, p2 represents the set of path nodes from the obstacle avoidance start point to the task end point, and m represents the number of path nodes. Indicates the cost of the new path. This represents the time cost of the i-th node. This represents the obstacle avoidance cost of the i-th node.

Images