Logistics robot cooperative control method and system for cargo sorting

Abstract

The invention relates to the technical field of robot control, and provides a logistics robot cooperative control method and system for cargo sorting. The method comprises the steps that a logistics robot in a cargo sorting bin is monitored in real time, and a monitoring block is obtained; predicting a robot fault risk based on the monitoring block, and determining a normal group and an abnormal group; according to a sorting priority rule, unexecuted tasks of the abnormal group are analyzed, and a complementary task sequence is obtained; collecting fixed task information of a normal group to obtain task distribution; optimizing normal group tasks based on task distribution and a complementary task sequence, and establishing an optimization space; performing optimization analysis according to the sorting evaluation model and the collaboration degree calculation network to obtain an optimization result; and performing cooperative control according to an optimization result. According to the method, the technical problems of low sorting efficiency and unreasonable task allocation caused by the fault risk of the logistics robot are solved, and the technical effects of optimizing sorting task allocation through cooperative control and improving the fault-tolerant capability and sorting efficiency of the robot are achieved.

Description

Technical Field

[0001] The present application relates to the field of robot control technology, and specifically to a collaborative control method and system for logistics robots used for cargo sorting. Background Art

[0002] With the rapid development of the e-commerce industry, the logistics industry is facing an increasing demand for cargo sorting. Traditional manual sorting methods can no longer meet the requirements of efficient, accurate and large-scale operations. Therefore, automated sorting systems have become one of the key technologies to improve logistics efficiency and accuracy. As an important part of the intelligent automated sorting system, logistics robots play an important role in improving sorting speed, reducing labor costs and improving operational safety. However, in the complex environment of multi-robot collaborative operations, logistics robots often face problems such as failure risks and uneven task allocation. Due to the large number of robots and the complexity and variety of tasks, how to accurately monitor the status of each robot, reasonably predict its failure risk, optimize task allocation and ensure that tasks are completed smoothly according to priority has become a technical problem that needs to be solved urgently. Most of the existing logistics robot control methods focus on the task optimization of a single robot, lack of in-depth consideration of the collaborative work of multiple robots, especially the response strategy when a robot fails, resulting in some robots being unable to perform tasks due to failure, thereby affecting the overall sorting efficiency. Summary of the invention

[0003] The present application provides a collaborative control method and system for a logistics robot for cargo sorting, aiming to solve the technical problems of reduced sorting efficiency and unreasonable task allocation due to the risk of failure of the logistics robot.

[0004] In view of the above problems, the present application provides a collaborative control method and system for logistics robots for cargo sorting.

[0005] In the first aspect disclosed in this application, a method for collaborative control of logistics robots for goods sorting is provided. The method includes: performing real-time monitoring on M logistics robots in a goods sorting warehouse to obtain M robot monitoring blocks, where M is a positive integer greater than 1; based on the M robot monitoring blocks, predicting the failure risks of the M logistics robots according to the robot failure risk prediction channel to determine a normal group of logistics robots and an abnormal group of logistics robots; according to the sorting priority evaluation rule, parsing the sorting priorities of the unexecuted sorting task set of the abnormal group of logistics robots to obtain a complementary sorting task sequence; collecting the fixed sorting task information of the normal group of logistics robots to obtain a fixed sorting task distribution; based on the fixed sorting task distribution, optimizing the sorting tasks of the normal group of logistics robots according to the complementary sorting task sequence to establish a sorting task optimization space; performing an optimization analysis on the sorting task optimization space according to the multi-dimensional sorting evaluation model and the sorting collaboration degree calculation network to obtain a sorting task optimization result; and performing sorting collaborative control of the abnormal group of logistics robots by the normal group of logistics robots according to the sorting task optimization result.

[0006] In another aspect disclosed in this application, a system for collaborative control of logistics robots for goods sorting is provided. The system includes: a real-time monitoring module: performing real-time monitoring on M logistics robots in a goods sorting warehouse to obtain M robot monitoring blocks, where M is a positive integer greater than 1; a failure risk prediction module: based on the M robot monitoring blocks, predicting the failure risks of the M logistics robots according to the robot failure risk prediction channel to determine a normal group of logistics robots and an abnormal group of logistics robots; a sorting priority parsing module: according to the sorting priority evaluation rule, parsing the sorting priorities of the unexecuted sorting task set of the abnormal group of logistics robots to obtain a complementary sorting task sequence; a task information collection module: collecting the fixed sorting task information of the normal group of logistics robots to obtain a fixed sorting task distribution; a sorting task optimization module: based on the fixed sorting task distribution, optimizing the sorting tasks of the normal group of logistics robots according to the complementary sorting task sequence to establish a sorting task optimization space; an optimization analysis module: performing an optimization analysis on the sorting task optimization space according to the multi-dimensional sorting evaluation model and the sorting collaboration degree calculation network to obtain a sorting task optimization result; and a collaborative control module: performing sorting collaborative control of the abnormal group of logistics robots by the normal group of logistics robots according to the sorting task optimization result.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages: The above-mentioned collaborative control method for logistics robots used in goods sorting first monitors M logistics robots in the warehouse in real time, and generates M robot monitoring blocks based on the monitoring results. Through these monitoring blocks, the working status of each robot can be captured in real time, providing data support for subsequent fault risk prediction. According to this data, the possible fault risks of each robot are evaluated using the fault risk prediction channel, and the robots are divided into a normal group and an abnormal group. Subsequently, according to the sorting priority evaluation rules, the priorities of the unexecuted sorting tasks in the abnormal group are analyzed. This process ensures that tasks with higher priorities can be completed first, and at the same time, through the analysis, a complementary sorting task sequence is generated, which will greatly improve the task execution efficiency and avoid task omission or conflict. On this basis, the fixed sorting task information of the robots in the normal group is further collected to obtain the fixed sorting task distribution. By analyzing this distribution, the sorting task load of the robots in the normal group can be accurately evaluated, and based on this information and the complementary task sequence, task optimization is carried out. The process of task optimization establishes a sorting task optimization space, and the establishment of this space helps to flexibly adjust task allocation in different scenarios to ensure that tasks are reasonably and evenly distributed among the robots. Finally, the sorting evaluation multi-dimensional model and the sorting cooperation degree calculation network are used to perform optimization analysis on the optimization space to maximize the cooperation degree of task sorting and generate an optimization result. This analysis process not only optimizes the task allocation but also improves the collaborative working efficiency of the robots. When the task optimization result is clear, the robots in the normal group will cooperate to execute the tasks of the robots in the abnormal group, thus ensuring the efficient operation of the entire sorting process.

[0008] The above description is only an overview of the technical solution of this application. In order to be able to understand the technical means of this application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features, and advantages of this application more obvious and understandable, the following specifically gives the specific implementation manners of this application. Brief Description of the Drawings

[0009] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0010] Figure 1 It is a schematic flowchart of a collaborative control method for logistics robots used in goods sorting in an embodiment.

[0011] Figure 2 It is an architecture diagram of a collaborative control system for logistics robots used in goods sorting in an embodiment.

[0012] Description of the drawing reference numerals: real-time monitoring module 1, fault risk prediction module 2, sorting priority analysis module 3, task information acquisition module 4, sorting task optimization module 5, optimization analysis module 6, collaborative control module 7. Specific embodiments

[0013] In an embodiment of the present application, there is provided a method and system for collaborative control of logistics robots for cargo sorting, which solves the technical problems of reduced sorting efficiency and unreasonable task allocation caused by the fault risk of logistics robots.

[0014] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the scope of protection of the present application.

[0015] It should be noted that the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or server including a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or are inherent to these processes, methods, products or devices.

[0016] Embodiment 1, as Figure 1 shown, the present application provides a method for collaborative control of logistics robots for cargo sorting, and the method includes: Performing real-time monitoring on M logistics robots in a cargo sorting warehouse to obtain M robot monitoring blocks, where M is a positive integer greater than 1.

[0017] In an embodiment of the present application, the system terminal performs real-time monitoring on multiple logistics robots (the number is M, and M is greater than 1) in the cargo sorting warehouse, and collects the status data of these logistics robots through sensors, cameras or other monitoring devices, such as motion status (speed, acceleration, etc.), battery status (voltage, current, temperature, etc.), motor status (voltage, current, temperature, etc.), etc.; the system terminal summarizes the status data of each logistics robot collected, and constructs M robot monitoring blocks, that is, a data set containing the operating conditions and other key performance indicators of the robot. In this way, the system terminal can grasp the working status of each robot in real time, ensure the accuracy and efficiency of the robot during the sorting process, and each robot monitoring block provides the necessary data support for subsequent fault prediction and task optimization, making the entire sorting process more intelligent and dynamically adjustable.

[0018] Based on the M robot monitoring blocks, the failure risk of the M logistics robots is predicted according to the robot failure risk prediction channel, and the normal group of logistics robots and the abnormal group of logistics robots are determined.

[0019] In one embodiment, based on the previously obtained M robot monitoring blocks, the system terminal will evaluate the failure risk of each robot through a failure risk prediction channel. Specifically, the system terminal will input the robot monitoring block corresponding to each robot into the robot failure risk prediction channel. The robot failure risk prediction channel predicts whether the corresponding robot will fail through the designated robot failure risk prediction branch inside, and assigns a risk coefficient to the corresponding robot, indicating the possibility of its failure. If the risk coefficient of a certain robot is relatively high, it will be classified into the abnormal group of logistics robots, indicating that the robot has a relatively high failure risk and needs special attention or suspension of tasks. If the failure risk is relatively low, the robot is classified into the normal group of logistics robots, indicating that its working state is stable and it can continue to execute tasks. Through this failure risk prediction mechanism, the system terminal can dynamically adjust the task allocation in the follow-up to ensure that the robots in the normal group continue to work efficiently, while the robots in the abnormal group are processed or repaired in a timely manner, thereby improving the reliability and efficiency of the entire sorting process.

[0020] Furthermore, the present application provides a method for predicting the failure risk of the M logistics robots according to the robot failure risk prediction channel based on the M robot monitoring blocks, and determining the normal group of logistics robots and the abnormal group of logistics robots, including: According to the M robot monitoring blocks, extract the mth robot monitoring block corresponding to the mth logistics robot, where m is a positive integer and 1 ≤ m ≤ M; based on the mth robot monitoring block, obtain the mth robot failure risk coefficient according to the robot failure risk prediction channel; determine whether the mth robot failure risk coefficient is less than the predetermined robot failure risk coefficient; if the mth robot failure risk coefficient is less than the predetermined robot failure risk coefficient, add the mth logistics robot to the normal group of logistics robots; if the mth robot failure risk coefficient is greater than or equal to the predetermined robot failure risk coefficient, add the mth logistics robot to the abnormal group of logistics robots.

[0021] Preferably, the system terminal first extracts the corresponding m-th robot monitoring block for each robot (numbered m, 1 ≤ m ≤ M) according to the M robot monitoring blocks obtained previously. This robot monitoring block contains the status data of the m-th robot, such as speed, battery temperature, motor temperature, etc.; subsequently, using the robot fault risk prediction channel, based on the data of the m-th robot's monitoring block, a fault risk prediction and analysis is carried out. The robot fault risk prediction channel transmits the status data of the m-th robot to the pre-trained prediction branch, thereby calculating the fault risk coefficient of the robot. The fault risk coefficient reflects the probability of the robot having a fault. The higher the value, the greater the risk of the robot having a fault; after that, the calculated fault risk coefficient of the m-th robot is compared with a predetermined robot fault risk coefficient, which is set according to historical data and robot operation experience, with the aim of determining whether the robot is within the normal working range. If the fault risk coefficient of the m-th robot is less than the predetermined robot fault risk coefficient, it indicates that the current working state of the robot is stable and the fault risk is low. The system terminal will add this robot to the normal group of logistics robots, indicating that the robot can continue to perform sorting tasks without intervention. If the fault risk coefficient of the m-th robot is greater than or equal to the predetermined robot fault risk coefficient, it indicates that the robot has a high fault risk and may have an abnormality or be at risk of an upcoming fault. The system terminal will add this robot to the abnormal group of logistics robots and may need to suspend its task for maintenance or take other remedial measures; the above grouping method helps the system terminal to manage the status of robots more effectively and provides support for subsequent task optimization and other operations, ensuring the efficient operation of the robots in the normal group, thereby improving the reliability and efficiency of the entire sorting process.

[0022] Furthermore, this application provides obtaining the m-th robot fault risk coefficient based on the m-th robot monitoring block according to the robot fault risk prediction channel, including: The robot fault risk prediction channel includes M robot fault risk prediction branches corresponding to the M logistics robots; based on the m-th logistics robot, activate the m-th robot fault risk prediction branch in the robot fault risk prediction channel, where the m-th robot fault risk prediction branch includes P robot fault risk prediction models corresponding to the m-th logistics robot, and P is a positive integer greater than 1; input the m-th robot monitoring block into the P robot fault risk prediction models to obtain P robot fault risk prediction coefficients; perform a mean calculation according to the P robot fault risk prediction coefficients to generate the m-th robot fault risk coefficient.

[0023] Optionally, the pre-built robot fault risk prediction channel includes M fault risk prediction branches corresponding to M logistics robots. Each robot (numbered m, 1 ≤ m ≤ M) has a dedicated prediction branch in the prediction channel. When it is necessary to predict the fault risk of the m-th robot, the system terminal will activate the fault risk prediction branch corresponding to the m-th robot, and this branch is responsible for performing risk assessment based on the monitoring data of this robot. The fault risk prediction branch of each robot contains P independent fault risk prediction models (where P is a positive integer greater than 1). These models are constructed based on different algorithms (such as machine learning models like neural networks and random forests). By predicting through multiple different algorithm models, the accuracy and reliability of the assessment of the robot's fault risk can be increased. The system terminal processes by taking the monitoring block of the m-th robot as the input and sending it into the P fault risk prediction models of the fault risk prediction branch of the m-th robot respectively. Each model can generate a fault risk prediction coefficient according to its specific prediction algorithm, indicating the assessment result of the model on the fault risk of the m-th robot. In order to integrate the assessment results of different models, the mean value of these P fault risk prediction coefficients is calculated to obtain the fault risk coefficient of the m-th robot. This mean value calculation method helps to improve the accuracy of the prediction and avoid the deviation of a single model from affecting the overall result.

[0024] For the construction of the robot failure risk prediction channel, the system terminal will build P failure risk prediction models based on neural networks, random forests, etc. By paralleling the P built failure risk prediction models, the m-th robot failure risk prediction branch is formed. Then, M robot failure risk prediction branches corresponding to M logistics robots are constructed in the same way, and these robot failure risk prediction branches are paralleled to form the robot failure risk prediction channel. Exemplarily, before building the P failure risk prediction models of the m-th robot failure risk prediction branch, the system terminal will collect the historical state data of the m-th robot, and use the expert decision-making method to label the risk coefficient for each piece of historical state data. Then, the labeled historical state data is divided into a training set and a test set. When using the multi-layer perceptron regression (MLP Regressor) in the neural network for construction, the system terminal will use the multi-layer perceptron regression to configure the failure risk prediction model, including the configuration of the input layer, hidden layer, output layer, hyperparameters (such as activation function, learning rate, maximum number of iterations, batch size, etc.). Subsequently, the weights of the failure risk prediction model are randomly initialized, usually using small random values. Then, the training set is input into the initialized failure risk prediction model. The risk prediction coefficient of the model is calculated through forward propagation. The mean squared error (MSE) is used to calculate the loss value between the risk prediction coefficient and the actual risk coefficient. The gradient of the loss value with respect to the model parameters is calculated through backpropagation, and the model weights are updated through the gradient descent algorithm to minimize the loss function. After the training reaches the maximum number of iterations, the system terminal uses the test set to evaluate the current failure risk prediction model. By calculating the mean squared error and coefficient of determination on the test set, it is judged whether the current model meets the expectations. If it meets the expectations, the current failure risk prediction model is output. If it does not meet the expectations, the model is optimized by adjusting the model structure (such as increasing the number of hidden layers, increasing the number of neurons, etc.) and adjusting the hyperparameter combination until it meets the expectations. When using the random forest regression in the random forest for construction, the system terminal will use the random forest regression to configure the failure risk prediction model, including the configuration of the number of trees, maximum depth, minimum number of samples for splitting, minimum number of leaf nodes, maximum number of features, etc. Then, the training set data is input into the configured failure risk prediction model. The model learns the features of the data by constructing multiple decision trees. Each tree randomly selects some features and samples on the training set to generate a subset of decision trees. During the training process, each tree classifies or regresses the data by splitting nodes until the specified stopping condition (such as maximum depth or minimum number of samples for splitting) is reached. Each tree gives a prediction value, and the random forest obtains the final regression result by averaging the prediction results of all trees. After the training ends, the model is also evaluated by calculating the mean squared error and coefficient of determination on the test set to judge whether the current model meets the expectations.

[0025] According to the sorting priority evaluation rules, the sorting priorities of the unexecuted sorting task set of the abnormal group of the logistics robots are analyzed to obtain a complementary sorting task sequence.

[0026] In one embodiment, the system terminal analyzes the priorities of the unexecuted sorting tasks in the abnormal group of the logistics robots according to the sorting priority evaluation rules, determines multi-dimensional indexes such as the urgency, importance, and complexity of each unexecuted sorting task in the unexecuted sorting task set, and then comprehensively evaluates the multi-dimensional indexes of each unexecuted sorting task to obtain the priority coefficient of each unexecuted sorting task. The tasks with higher priority coefficients are regarded as tasks to be processed preferentially. All the tasks in the unexecuted sorting task set are sorted from high to low according to these priority coefficients to generate a complementary sorting task sequence. This sequence ensures that when the robot fails or is unable to execute tasks, the tasks in the abnormal group can be quickly replaced and completed, ensuring that no tasks are missed due to robot failures, and at the same time maximizing the utilization of resources.

[0027] Furthermore, the present application provides a method for parsing the sorting priorities of the unexecuted sorting task set of the abnormal group of the logistics robots according to the sorting priority evaluation rules to obtain a complementary sorting task sequence, including: The sorting priority evaluation rules include multi-dimensional indexes for sorting priority evaluation and multi-dimensional weights for sorting priority evaluation. Among them, the multi-dimensional indexes for sorting priority evaluation include the urgency of the sorting task, the importance of the sorting task, and the complexity of the sorting task, and the multi-dimensional weights for sorting priority evaluation include the weight of the urgency of the sorting task, the weight of the importance of the sorting task, and the weight of the complexity of the sorting task; according to the multi-dimensional indexes for sorting priority evaluation, the sorting priority evaluations of the unexecuted sorting task set are carried out to obtain the sorting priority evaluation sets of each task; according to the multi-dimensional weights for sorting priority evaluation, weighted calculations are respectively carried out on the sorting priority evaluation sets of each task to obtain the sorting priority coefficients of each task; according to the sorting priority coefficients of each task, the unexecuted sorting task set is sorted in terms of sorting priority to generate the complementary sorting task sequence.

[0028] Preferably, the sorting priority evaluation rule includes two main parts, namely the multi-dimensional index of sorting priority evaluation and the multi-dimensional weight of sorting priority evaluation. The multi-dimensional index of sorting priority evaluation is used to measure the priority of each task, including the urgency of the sorting task, the importance of the sorting task, and the complexity of the sorting task. Among them, the urgency of the sorting task refers to the timeliness requirement for task completion. The higher the urgency of the task, the higher the priority. For example, a task that must be completed in a short time has a higher urgency. The importance of the sorting task refers to the importance of the task. The greater the impact of the task on subsequent tasks, the higher the importance of the task. For example, a sorting task involving important goods usually has a higher importance. The complexity of the sorting task refers to the difficulty of the task or the complexity of the operation. A task with a high complexity may require more resources and time, so it may be ranked relatively low in the priority sorting. The multi-dimensional weight of sorting priority evaluation includes the weight of the urgency of the sorting task, the weight of the importance of the sorting task, and the weight of the complexity of the sorting task, which reflects the relative importance of each index in the overall priority evaluation and is determined according to business requirements and expert decision-making methods; for each unexecuted sorting task in the unexecuted sorting task set, the system terminal will evaluate according to the multi-dimensional index of sorting priority evaluation. Specifically, when evaluating the urgency of the sorting task, the system terminal will calculate the difference between the scheduled completion time of the task and the current time to obtain the remaining time of the task, and then use Quantify the urgency of the sorting tasks; when evaluating the importance of the sorting tasks, the system terminal calculates the ratio of the value of the task to the maximum task value to obtain the value impact coefficient of the task, and then multiplies this value impact coefficient by the task type weight to quantify the importance of the sorting tasks. Among them, the task type weight is determined based on historical experience and expert decision-making. For example, the task type weight of urgent orders is 1, and the task type weight of ordinary orders is 0.5, etc.; when evaluating the complexity of the sorting tasks, the system terminal will obtain the sorting steps involved in the task, such as picking up goods, sorting, handling, etc., and then obtain the number of robots required for each step. By multiplying the number of sorting steps by the number of robots and dividing the product by the standard complexity coefficient, the step complexity is obtained. Subsequently, the ratio of the scheduled completion time of the task to the standard task time is calculated to obtain the time complexity. By adding the step complexity and the time complexity, the sorting task complexity is obtained; after the sorting priority evaluation of each unexecuted sorting task is completed, the system terminal summarizes the sorting task urgency, sorting task importance, and sorting task complexity of each unexecuted sorting task to form a sorting priority evaluation set for each task; then, the sorting task urgency, sorting task importance, and sorting task complexity in the sorting priority evaluation set for each task are weighted and summed using the multi-dimensional weights of the sorting priority evaluation to obtain the sorting priority coefficient for each task; then, the unexecuted sorting task set is sorted from high to low according to the sorting priority coefficient for each task to generate a complementary sorting task sequence, ensuring that the most important tasks are not missed during task allocation and execution, and the execution order between each task can be balanced to avoid waste of resources.

[0029] Collect the fixed sorting task information of the normal group of the logistics robot to obtain the fixed sorting task distribution.

[0030] In one embodiment, the system terminal collects the fixed sorting task information that has been assigned to the normal group of the logistics robot through the interface with the sorting task management system. This fixed sorting task information refers to those sorting tasks that have been assigned to the robot but have not been executed, including the specific information of each task. For example, the type of the task (picking up goods, sorting, handling, etc.), the scheduled execution time, the scheduled completion time, the task priority, etc.; according to the collected fixed sorting task information, the system terminal determines the distribution of these tasks among different robots. The task distribution not only involves the number of tasks for each robot but also includes information such as the specific execution order of the tasks and the task type. This process can help the system terminal understand the current task load distribution situation, ensure the balanced distribution of tasks, and avoid some robots being overloaded or idle, thereby improving the overall sorting efficiency.

[0031] Based on the fixed sorting task distribution, the sorting tasks of the normal group of logistics robots are optimized according to the complementary sorting task sequence, and a sorting task optimization space is established.

[0032] In one embodiment, the system terminal optimizes the sorting tasks of the normal group of logistics robots according to the fixed sorting task distribution, that is, the information of the tasks that have been assigned but not executed, combined with the complementary sorting task sequence, so as to establish a sorting task optimization space; specifically, the system terminal first counts the current task volume undertaken by each robot in the normal group of logistics robots through the fixed sorting task distribution, and screens the robots that can continue to undertake the tasks; then, tasks are selected from the complementary sorting task sequence and randomly inserted into the robots that are not fully loaded. These tasks are arranged according to priority, and according to the current load of the robot, they are preferentially allocated to those robots that are far from the load limit (that is, the difference between the maximum task volume that can be undertaken and the current task volume that has been undertaken is large). In this way, the system terminal completes the preliminary sorting task optimization and forms a sorting task optimization decision; repeating the above process, the system terminal generates multiple sorting task optimization decisions, and establishes a sorting task optimization space based on these sorting task optimization decisions. This space contains multiple sorting task optimization decisions, each of which takes into account the load situation and task priority of the robot, providing data support for subsequent optimization analysis.

[0033] The optimization space of the sorting task is optimized and analyzed according to the sorting evaluation multidimensional model and the sorting coordination degree calculation network to obtain the sorting task optimization result.

[0034] In one embodiment, the system terminal further optimizes and analyzes the sorting task optimization space by using the sorting evaluation multidimensional model and the sorting coordination calculation network, with the aim of finding the optimal task allocation plan, thereby improving the efficiency of the entire sorting; specifically, the system terminal analyzes and calculates the sorting task optimization decisions in the sorting task optimization space through the collaborative processing of the sorting evaluation multidimensional model and the sorting coordination calculation network, and generates the sorting coordination of the sorting task optimization decisions. The system terminal will automatically select an optimal sorting task optimization decision from multiple decisions as the sorting task optimization result according to the coordination of each decision. This optimal decision ensures the efficient use of resources and the smooth completion of tasks, thereby maximizing the overall efficiency of the sorting process.

[0035] Furthermore, the present application provides a method for performing an optimization analysis on the optimization space of the sorting task according to the sorting evaluation multidimensional model and the sorting coordination degree calculation network to obtain a sorting task optimization result, including: Based on the sorting evaluation multi-dimensional model and the sorting cooperation degree calculation network, the sorting cooperation degree of the sorting task optimization space is analyzed to obtain multiple sorting cooperation degrees; it is judged whether the multiple sorting cooperation degrees are greater than or equal to a predetermined sorting cooperation degree to obtain multiple sorting cooperation degree judgment results; based on the multiple sorting cooperation degree judgment results, the sorting task optimization space is optimized and screened to establish a sorting task optimization decision domain; according to the sorting task optimization decision domain, the maximum optimization of the sorting cooperation degree is carried out to generate the sorting task optimization result.

[0036] Preferably, the system terminal first uses the sorting evaluation multi-dimensional model and the sorting cooperation degree calculation network to analyze multiple sorting task optimization decisions in the sorting task optimization space. Each sorting task optimization decision will be analyzed and processed by the sorting evaluation multi-dimensional model to generate a corresponding sorting evaluation result. The sorting evaluation result of each sorting task optimization decision will also be comprehensively calculated by the sorting cooperation degree calculation network to generate a corresponding sorting cooperation degree. The sorting cooperation degree is a key indicator to measure the coordination between robots and the task execution efficiency in the sorting task optimization decision. Subsequently, the system terminal compares the multiple sorting cooperation degrees parsed with a predetermined sorting cooperation degree (determined based on business requirements and expert decisions), and judges whether each sorting cooperation degree reaches or exceeds the predetermined sorting cooperation degree. If the sorting cooperation degree of a certain sorting task optimization decision is greater than or equal to the predetermined sorting cooperation degree, it is considered that the decision meets the basic requirements. On the contrary, if the sorting cooperation degree of a certain sorting task optimization decision is less than the predetermined sorting cooperation degree, it is considered that the decision does not meet the requirements and is excluded from the optimization process. After comparing the judgment results, all eligible sorting task optimization decisions (i.e., the solutions with sorting cooperation degrees greater than or equal to the predetermined sorting cooperation degree) are added to the sorting task optimization decision domain. The sorting task optimization decision domain is a set composed of sorting task optimization decisions that meet the predetermined cooperation degree standard, and this set is used for further optimization. Based on the sorting task optimization decision domain, the system terminal performs the maximum optimization of the sorting cooperation degree for the sorting task optimization decisions in the sorting task optimization decision domain. The goal of this step is to select the task allocation plan with the highest sorting cooperation degree from the optimization decision domain to ensure that the efficiency and coordination of multi-robot task allocation reach the optimal. The system terminal can screen out the decision with the highest sorting cooperation degree from the sorting task optimization decision domain through the maximum optimization as the final sorting task optimization result. This result is the optimal sorting task optimization decision in the current sorting task optimization space, which can achieve the double improvement of sorting efficiency and resource utilization rate, and at the same time ensure the best task coordination between robots.

[0037] Furthermore, the present application provides an analysis of the sorting cooperation degree of the sorting task optimization space based on the sorting evaluation multi-dimensional model and the sorting cooperation degree calculation network to obtain multiple sorting cooperation degrees, including: Traverse the sorting task optimization space to extract the first sorting task optimization decision; input the first sorting task optimization decision into the sorting evaluation multi-dimensional model to obtain the first sorting evaluation result; determine whether the first sorting evaluation result meets the sorting evaluation multi-dimensional constraints, where the sorting evaluation multi-dimensional constraints include sorting efficiency constraints, sorting task allocation balance constraints, and sorting resource utilization rate constraints; if the first sorting evaluation result does not meet the sorting evaluation multi-dimensional constraints, output the first sorting cooperation degree as 0; if the first sorting evaluation result meets the sorting evaluation multi-dimensional constraints, activate the sorting cooperation degree calculation network, where the sorting cooperation degree calculation network includes sorting efficiency weights, sorting task allocation balance weights, and sorting resource utilization rate weights; input the first sorting evaluation result into the sorting cooperation degree calculation network to obtain the first sorting cooperation degree, and add the first sorting cooperation degree to the multiple sorting cooperation degrees.

[0038] Optionally, the system terminal randomly extracts a sorting task optimization decision from the sorting task optimization space as the first sorting task optimization decision, and each decision represents a possible task allocation combination; subsequently, the first sorting task optimization decision is input into a pre-constructed sorting evaluation multi-dimensional model, which will comprehensively evaluate the first sorting task optimization decision through an internal sorting efficiency evaluation model, a sorting task allocation balance evaluation model, and a sorting resource utilization evaluation model, generating a multi-dimensional evaluation result, including sorting efficiency (the speed and time of task completion), task allocation balance (the balance of the task loads of each robot), and resource utilization (the usage efficiency of the sorting task for system resources such as robots and power); then, the system terminal compares the first sorting evaluation result with a predetermined sorting evaluation multi-dimensional constraint to determine whether it meets the sorting efficiency constraint, the sorting task allocation balance constraint, and the sorting resource utilization constraint in the sorting evaluation multi-dimensional constraint. This sorting evaluation multi-dimensional constraint is also determined based on business requirements and expert decisions and is used to evaluate whether the completion efficiency of the sorting task optimization decision meets the expectations, whether the task allocation avoids overloading or robot idleness, and whether the resource usage is efficient and without waste. If the first sorting evaluation result does not meet the above constraint conditions, the system terminal will set the first sorting coordination degree of the first sorting task optimization decision to 0, and this decision will be considered invalid and will not enter the further coordination degree calculation. If the first sorting evaluation result meets the multi-dimensional constraint conditions, the system terminal will activate the sorting coordination degree calculation network to further evaluate the coordination degree of this scheme. The sorting coordination degree calculation network internally contains multiple preset weights, namely the sorting efficiency weight, the sorting task allocation balance weight, and the sorting resource utilization weight, which are used to measure the influence of the corresponding evaluation indicators on the coordination degree and are all determined through business requirements and expert decisions; by inputting the first sorting evaluation result into the sorting coordination degree calculation network, the system terminal can calculate the first sorting coordination degree of the first sorting task optimization decision in a weighted summation manner according to the evaluation result and the weights of each evaluation indicator. This first sorting coordination degree will be added to multiple sorting coordination degrees to facilitate subsequent comparison and optimization of multiple sorting task optimization decisions, ensuring the efficiency, balance, and resource utilization of the sorting task allocation.

[0039] Furthermore, the present application provides inputting the first sorting task optimization decision into the sorting evaluation multi-dimensional model to obtain a first sorting evaluation result, including: The sorting evaluation multi-dimensional model includes a sorting efficiency evaluation model, a sorting task allocation balance evaluation model, and a sorting resource utilization rate evaluation model; input the first sorting task optimization decision into the sorting efficiency evaluation model to obtain the first sorting efficiency; input the first sorting task optimization decision into the sorting task allocation balance evaluation model to obtain the first sorting task allocation balance; input the first sorting task optimization decision into the sorting resource utilization rate evaluation model to output the first sorting resource utilization rate; output the first sorting efficiency, the first sorting task allocation balance, and the first sorting resource utilization rate as the first sorting evaluation result.

[0040] Optionally, the pre-constructed sorting evaluation multi-dimensional model includes a sorting efficiency evaluation model, a sorting task allocation balance evaluation model, and a sorting resource utilization rate evaluation model. The sorting efficiency evaluation model is used to evaluate the sorting efficiency of the task allocation scheme. The sorting task allocation balance evaluation model is used to measure the balance of task allocation to ensure load balance among robots. The sorting resource utilization rate evaluation model is used to evaluate whether the resource usage is efficient. The construction methods of these models are the same as those of the aforementioned construction of the robot failure risk prediction branch, and can all be constructed through methods such as neural networks and random forests. For example, if a multi-layer perceptron regression in a neural network is used, it is also trained through steps such as forward propagation, loss calculation, backward propagation, and parameter optimization to construct these evaluation models. Then, by paralleling these evaluation models, a sorting evaluation multi-dimensional model is constructed; the system terminal inputs the first sorting task optimization decision into the sorting evaluation multi-dimensional model. The sorting evaluation multi-dimensional model respectively transmits the first sorting task optimization decision to the internal evaluation models. Through the sorting efficiency evaluation model, calculate the first sorting efficiency of the first sorting task optimization decision. Through the sorting task allocation balance evaluation model, calculate the first task allocation balance of the first sorting task optimization decision. Through the sorting resource utilization rate evaluation model, calculate the first sorting resource utilization rate of the first sorting task optimization decision. Then, integrate the results of the first sorting efficiency, the first sorting task allocation balance, and the first sorting resource utilization rate into a multi-dimensional first sorting evaluation result, providing an important basis for subsequent sorting coordination calculation and optimization decision-making.

[0041] Perform sorting collaborative control of the abnormal group of logistics robots according to the sorting task optimization result and the normal group of logistics robots.

[0042] In one embodiment, after obtaining the optimized result of the sorting task, the system terminal will allocate tasks to the normal group of logistics robots according to the optimized result, and coordinate the execution of the sorting tasks originally belonging to the abnormal group of logistics robots to ensure that the tasks of the abnormal group are processed in a timely manner. During the task execution process, the system terminal will monitor the completion status of the tasks in real time and continuously update the task queue. If a task is completed or a robot malfunctions, the system terminal will promptly adjust the allocation of the remaining tasks. Finally, through the collaborative work of the robots in the normal group, not only can their own tasks be completed, but also the uncompleted tasks of the abnormal group can be supplemented and completed, ensuring that the entire sorting task is completed as planned, improving the overall sorting efficiency and stability.

[0043] In summary, the embodiments of the present application have at least the following technical effects: Based on the real-time monitoring of M logistics robots, the embodiments of the present application classify the robots into a normal group and an abnormal group through a robot failure risk prediction channel, and analyze the priorities of the uncompleted tasks of the abnormal group to generate a complementary sorting task sequence; combined with the fixed task information of the normal group, the task allocation scheme is optimized, and the sorting task optimization space is gradually constructed; during the optimization process, a multi-dimensional sorting evaluation model and a cooperation degree calculation network are applied to analyze the task allocation decisions in the optimization space from multiple dimensions such as sorting efficiency, task balance degree, and resource utilization rate; by screening the decisions with high cooperation degrees, an optimized decision domain is established, and further optimization is carried out on this basis to generate the sorting task optimization result with the highest cooperation degree; finally, according to the optimized result, the normal group is coordinated to complete the tasks of the abnormal group, realizing task load balance and efficient cooperation; these technical effects jointly solve the technical problems of reduced sorting efficiency and unreasonable task allocation caused by the failure risk of logistics robots, and achieve the technical effects of optimizing the sorting task allocation through collaborative control, improving the robot failure tolerance ability and sorting efficiency.

[0044] Embodiment 2, based on the same inventive concept as the method for collaborative control of logistics robots for cargo sorting in the foregoing embodiment, as Figure 2As shown, the present application provides a collaborative control system for logistics robots used in goods sorting. The system includes: Real-time monitoring module 1: Real-time monitor M logistics robots in the goods sorting warehouse to obtain M robot monitoring blocks, where M is a positive integer greater than 1; Fault risk prediction module 2: Based on the M robot monitoring blocks, predict the fault risks of the M logistics robots according to the robot fault risk prediction channel, and determine a normal group of logistics robots and an abnormal group of logistics robots; Sorting priority analysis module 3: According to the sorting priority evaluation rule, analyze the sorting priorities of the unexecuted sorting task sets of the abnormal group of logistics robots to obtain a complementary sorting task sequence; Task information collection module 4: Collect the fixed sorting task information of the normal group of logistics robots to obtain the fixed sorting task distribution; Sorting task optimization module 5: Based on the fixed sorting task distribution, optimize the sorting tasks of the normal group of logistics robots according to the complementary sorting task sequence to establish a sorting task optimization space; Optimization analysis module 6: Perform optimization analysis on the sorting task optimization space according to the sorting evaluation multi-dimensional model and the sorting collaboration degree calculation network to obtain a sorting task optimization result; Collaborative control module 7: Perform sorting collaborative control on the normal group of logistics robots according to the sorting task optimization result and the abnormal group of logistics robots.

[0045] Further, the fault risk prediction module 2 is further configured to execute the following method: According to the M robot monitoring blocks, extract the mth robot monitoring block corresponding to the mth logistics robot, where m is a positive integer and 1 ≤ m ≤ M; Based on the mth robot monitoring block, obtain the mth robot fault risk coefficient according to the robot fault risk prediction channel; Determine whether the mth robot fault risk coefficient is less than a predetermined robot fault risk coefficient; If the mth robot fault risk coefficient is less than the predetermined robot fault risk coefficient, add the mth logistics robot to the normal group of logistics robots; If the mth robot fault risk coefficient is greater than or equal to the predetermined robot fault risk coefficient, add the mth logistics robot to the abnormal group of logistics robots.

[0046] Further, the fault risk prediction module 2 is further configured to execute the following method: The robot fault risk prediction channel includes M robot fault risk prediction branches corresponding to the M logistics robots; based on the m-th logistics robot, the m-th robot fault risk prediction branch in the robot fault risk prediction channel is activated, where the m-th robot fault risk prediction branch includes P robot fault risk prediction models corresponding to the m-th logistics robot, and P is a positive integer greater than 1; the m-th robot monitoring block is input into the P robot fault risk prediction models to obtain P robot fault risk prediction coefficients; the mean value is calculated according to the P robot fault risk prediction coefficients to generate the m-th robot fault risk coefficient.

[0047] Further, the sorting priority analysis module 3 is further configured to execute the following method: The sorting priority evaluation rule includes sorting priority evaluation multi-dimensional indexes and sorting priority evaluation multi-dimensional weights. Among them, the sorting priority evaluation multi-dimensional indexes include the urgency of the sorting task, the importance of the sorting task, and the complexity of the sorting task. The sorting priority evaluation multi-dimensional weights include the urgency weight of the sorting task, the importance weight of the sorting task, and the complexity weight of the sorting task; according to the sorting priority evaluation multi-dimensional indexes, the sorting priority of the unexecuted sorting task set is evaluated to obtain the sorting priority evaluation set of each task; according to the sorting priority evaluation multi-dimensional weights, weighted calculation is respectively performed on the sorting priority evaluation sets of each task to obtain the sorting priority coefficient of each task; according to the sorting priority coefficients of each task, the unexecuted sorting task set is sorted by sorting priority to generate the complementary sorting task sequence.

[0048] Further, the optimization analysis module 6 is further configured to execute the following method: Based on the sorting evaluation multi-dimensional model and the sorting cooperation degree calculation network, the sorting cooperation degree of the sorting task optimization space is analyzed to obtain multiple sorting cooperation degrees; it is judged whether the multiple sorting cooperation degrees are greater than or equal to a predetermined sorting cooperation degree to obtain multiple sorting cooperation degree judgment results; based on the multiple sorting cooperation degree judgment results, the sorting task optimization space is optimized and screened to establish a sorting task optimization decision domain; according to the sorting task optimization decision domain, the maximum optimization of the sorting cooperation degree is performed to generate the sorting task optimization result.

[0049] Further, the optimization analysis module 6 is further configured to execute the following method: Traverse the sorting task optimization space, and extract the first sorting task optimization decision; input the first sorting task optimization decision into the multi-dimensional sorting evaluation model to obtain the first sorting evaluation result; determine whether the first sorting evaluation result meets the multi-dimensional sorting evaluation constraints, where the multi-dimensional sorting evaluation constraints include sorting efficiency constraints, sorting task allocation balance constraints, and sorting resource utilization rate constraints; if the first sorting evaluation result does not meet the multi-dimensional sorting evaluation constraints, output the first sorting collaboration degree as 0; if the first sorting evaluation result meets the multi-dimensional sorting evaluation constraints, activate the sorting collaboration degree calculation network, where the sorting collaboration degree calculation network includes sorting efficiency weights, sorting task allocation balance weights, and sorting resource utilization rate weights; input the first sorting evaluation result into the sorting collaboration degree calculation network to obtain the first sorting collaboration degree, and add the first sorting collaboration degree to the multiple sorting collaboration degrees.

[0050] Further, the optimization analysis module 6 is further configured to execute the following method: The multi-dimensional sorting evaluation model includes a sorting efficiency evaluation model, a sorting task allocation balance evaluation model, and a sorting resource utilization rate evaluation model; input the first sorting task optimization decision into the sorting efficiency evaluation model to obtain the first sorting efficiency; input the first sorting task optimization decision into the sorting task allocation balance evaluation model to obtain the first sorting task allocation balance; input the first sorting task optimization decision into the sorting resource utilization rate evaluation model to output the first sorting resource utilization rate; output the first sorting efficiency, the first sorting task allocation balance, and the first sorting resource utilization rate as the first sorting evaluation result.

[0051] It should be noted that the above sequence of embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above description of specific embodiments of this specification has been made. The processes depicted in the drawings do not necessarily require the specific order and continuous order shown to achieve the desired result. In some embodiments, multi-task processing and parallel processing are also possible or may be advantageous.

[0052] The above are only the preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

[0053] This specification and the drawings are merely exemplary illustrations of the present application and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of the present application. Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application is intended to include these changes and modifications therein.

Claims

1. A collaborative control method for logistics robots used for cargo sorting, characterized in that: The method comprises: Perform real-time monitoring on M logistics robots in the cargo sorting warehouse to obtain M robot monitoring blocks, where M is a positive integer greater than 1; Based on the M robot monitoring blocks, a fault risk prediction is performed on the M logistics robots according to a robot fault risk prediction channel to determine a normal group of logistics robots and an abnormal group of logistics robots; According to the sorting priority evaluation rule, the unexecuted sorting task set of the abnormal group of logistics robots is parsed according to the sorting priority to obtain a complementary sorting task sequence; Collecting fixed sorting task information of the normal group of logistics robots to obtain fixed sorting task distribution; Based on the fixed sorting task distribution, the sorting tasks of the normal group of logistics robots are optimized according to the complementary sorting task sequence to establish a sorting task optimization space; Performing optimization analysis on the optimization space of the sorting task according to the sorting evaluation multidimensional model and the sorting coordination degree calculation network to obtain the sorting task optimization result; Sorting collaborative control of the abnormal group of logistics robots is performed according to the sorting task optimization result and the normal group of logistics robots.

2. The method according to claim 1, characterized in that According to the sorting priority evaluation rule, the unexecuted sorting task set of the abnormal group of logistics robots is parsed according to the sorting priority to obtain a complementary sorting task sequence, including: The sorting priority evaluation rule includes a sorting priority evaluation multidimensional index and a sorting priority evaluation multidimensional weight, wherein the sorting priority evaluation multidimensional index includes the urgency of the sorting task, the importance of the sorting task and the complexity of the sorting task, and the sorting priority evaluation multidimensional weight includes the urgency weight of the sorting task, the importance weight of the sorting task and the complexity weight of the sorting task; According to the multi-dimensional index of sorting priority evaluation, the sorting priority of the unexecuted sorting task set is evaluated to obtain a sorting priority evaluation set of each task; According to the multi-dimensional weight of the sorting priority evaluation, weighted calculation is performed on each task sorting priority evaluation set to obtain a sorting priority coefficient for each task; According to the sorting priority coefficients of each task, the unexecuted sorting task set is sorted in priority order to generate the complementary sorting task sequence.

3. The method according to claim 1, characterized in that The optimization space of the sorting task is optimized and analyzed according to the sorting evaluation multidimensional model and the sorting coordination degree calculation network to obtain the sorting task optimization result, including: Based on the sorting evaluation multidimensional model and the sorting coordination degree calculation network, the sorting task optimization space is analyzed for sorting coordination degree to obtain multiple sorting coordination degrees; Determining whether the plurality of sorting coordination degrees are greater than or equal to a predetermined sorting coordination degree, and obtaining a plurality of sorting coordination degree determination results; Based on the multiple sorting coordination degree judgment results, the sorting task optimization space is optimized and screened to establish a sorting task optimization decision domain; The sorting coordination degree is maximized and optimized according to the sorting task optimization decision domain to generate the sorting task optimization result.

4. The method according to claim 3, characterized in that Based on the sorting evaluation multidimensional model and the sorting coordination degree calculation network, the sorting task optimization space is analyzed for sorting coordination degree to obtain multiple sorting coordination degrees, including: Traversing the sorting task optimization space and extracting the first sorting task optimization decision; Inputting the first sorting task optimization decision into the sorting evaluation multidimensional model to obtain a first sorting evaluation result; Determining whether the first sorting evaluation result satisfies the sorting evaluation multidimensional constraints, wherein the sorting evaluation multidimensional constraints include sorting efficiency constraints, sorting task allocation balance constraints, and sorting resource utilization constraints; If the first sorting evaluation result does not satisfy the multi-dimensional constraint of the sorting evaluation, outputting the first sorting coordination degree as 0; If the first sorting evaluation result satisfies the sorting evaluation multidimensional constraint, activating the sorting coordination degree calculation network, wherein the sorting coordination degree calculation network includes a sorting efficiency weight, a sorting task allocation balance weight, and a sorting resource utilization weight; Inputting the first sorting evaluation result into the sorting coordination degree calculation network to obtain the first sorting coordination degree, The first sorting coordination degree is added to the plurality of sorting coordination degrees.

5. The method according to claim 4, characterized in that Inputting the first sorting task optimization decision into the sorting evaluation multidimensional model to obtain a first sorting evaluation result includes: The sorting evaluation multidimensional model includes a sorting efficiency evaluation model, a sorting task allocation balance evaluation model and a sorting resource utilization evaluation model; Inputting the first sorting task optimization decision into the sorting efficiency evaluation model to obtain a first sorting efficiency; Inputting the first sorting task optimization decision into the sorting task allocation balance evaluation model to obtain the first sorting task allocation balance; Inputting the first sorting task optimization decision into the sorting resource utilization evaluation model, and outputting the first sorting resource utilization; The first sorting efficiency, the first sorting task allocation balance and the first sorting resource utilization rate are output as the first sorting evaluation result.

6. The method according to claim 1, characterized in that Based on the M robot monitoring blocks, the M logistics robots are predicted for failure risks according to the robot failure risk prediction channel, and a normal group of logistics robots and an abnormal group of logistics robots are determined, including: According to the M robot monitoring blocks, extract the mth robot monitoring block corresponding to the mth logistics robot, where m is a positive integer, 1≤m≤M; Based on the mth robot monitoring block, and according to the robot failure risk prediction channel, obtaining the mth robot failure risk coefficient; Determining whether the mth robot failure risk coefficient is less than a predetermined robot failure risk coefficient; If the m-th robot failure risk coefficient is less than the predetermined robot failure risk coefficient, adding the m-th logistics robot to the normal group of logistics robots; If the m-th robot failure risk coefficient is greater than or equal to the predetermined robot failure risk coefficient, the m-th logistics robot is added to the logistics robot abnormal group.

7. The method according to claim 6, characterized in that Based on the mth robot monitoring block and according to the robot fault risk prediction channel, obtaining the mth robot fault risk coefficient includes: The robot failure risk prediction channel includes M robot failure risk prediction branches corresponding to the M logistics robots; Based on the mth logistics robot, activating the mth robot fault risk prediction branch in the robot fault risk prediction channel, wherein the mth robot fault risk prediction branch includes P robot fault risk prediction models corresponding to the mth logistics robot, where P is a positive integer greater than 1; Inputting the mth robot monitoring block into the P robot failure risk prediction models to obtain P robot failure risk prediction coefficients; The mean value of the P robot failure risk prediction coefficients is calculated to generate the m-th robot failure risk coefficient.

8. A logistics robot collaborative control system for cargo sorting, characterized in that: The steps for implementing the collaborative control method of a logistics robot for cargo sorting according to any one of claims 1 to 7 include: Real-time monitoring module: real-time monitoring of M logistics robots in the cargo sorting warehouse to obtain M robot monitoring blocks, where M is a positive integer greater than 1; Fault risk prediction module: based on the M robot monitoring blocks, the fault risk prediction module performs fault risk prediction on the M logistics robots according to the robot fault risk prediction channel, and determines a normal group of logistics robots and an abnormal group of logistics robots; Sorting priority parsing module: according to the sorting priority evaluation rules, the unexecuted sorting task set of the abnormal group of logistics robots is parsed to obtain a complementary sorting task sequence; Task information collection module: collects the fixed sorting task information of the normal group of logistics robots to obtain the fixed sorting task distribution; Sorting task optimization module: based on the fixed sorting task distribution, the sorting task of the normal group of logistics robots is optimized according to the complementary sorting task sequence, and a sorting task optimization space is established; Optimization analysis module: performs optimization analysis on the optimization space of the sorting task according to the sorting evaluation multidimensional model and the sorting coordination degree calculation network to obtain the sorting task optimization result; Collaborative control module: performs sorting collaborative control of the abnormal group of logistics robots according to the sorting task optimization result and the normal group of logistics robots.

Images