A task management method and system for a loading robot

Abstract

The present application belongs to the field of logistics automation, and specifically discloses a task management method and system for a loading robot, the method comprising: collecting current date data, and using an LSTM model to predict the loading volume data of the day according to the date data; judging whether a user has issued a loading task instruction, collecting current environmental data according to the loading task instruction, and extracting loading task information; obtaining a task execution plan according to the loading volume data, environmental data, and loading task information using a trained deep learning model; determining that a truck is in place, and adjusting the loading robot to perform loading according to the task execution plan; determining that the loading robot has completed the loading work of the day, obtaining the actual loading volume data of the day to correct the LSTM model; obtaining all loading information corresponding to the loading task of the day, and saving the loading information as historical data to a database. The present application can improve the efficiency of the loading process.

Description

Technical Field

[0001] The present application belongs to the field of logistics automation, and more specifically, to a task management method and system for a loading robot. Background Art

[0002] Cargo loading is an important part of the modern logistics system. The goal is to rely on transport vehicles to carry out the transportation of goods between warehouses, production lines, distribution points and other logistics areas that are dispersed in location, so that long-distance and large-scale logistics and various stages of production can be effectively connected through the transportation of goods, and large-scale material conversion and processing can be achieved. Loading has become an indispensable part of large-scale production and logistics systems in industries such as manufacturing, chemical industry, and agriculture.

[0003] However, the current method of loading and unloading goods by transport vehicles is still mainly manual handling, which is inefficient and slow. The management method cannot meet the needs of large-scale loading and unloading of modern goods, especially it is easy to cause safety risks for loading and unloading personnel; and the cost of manual loading is also very high. The labor costs of some large enterprises for loading are as high as tens of millions of yuan each year, especially the loading costs of bagged goods can account for more than one-third of the total transportation cost.

[0004] Furthermore, in the research of the above-mentioned problems in the prior art, there are still problems such as long working hours, large loading and unloading volumes, and many types of goods for loading robots. However, the current loading task management method is simple, and generally only determines the loading plan based on simple factors such as the amount of goods and task sequencing, but lacks intelligent analysis of more complex factors such as the environment and goods characteristics, which can easily lead to long loading and unloading time, easy damage of goods, unreasonable loading and unloading, and affect the life of the robot, thereby reducing the utilization efficiency of the loading robot.

[0005] Therefore, how to conduct intelligent analysis of the multiple complex factors of loading tasks to improve loading quality, efficiency and service life of loading robots is a technical problem that urgently needs to be solved. Summary of the invention

[0006] In view of the defects of the prior art, the purpose of this application is to provide a task management method and system for a loading robot, aiming to solve the problems in the prior art of the loading robot such as long loading and unloading time, easy damage of goods, unreasonable loading and unloading, and affected robot life.

[0007] To achieve the above objectives, in a first aspect, the present application provides a task management method for a loading robot, comprising:

[0008] Collect current date data, and use the LSTM model to predict the loading volume data of the day based on the date data;

[0009] Determine whether a user has issued a loading task instruction, collect current environmental data according to the loading task instruction, and extract loading task information;

[0010] Obtain the task execution plan using the trained deep learning model based on the loading volume data, environmental data, and loading task information;

[0011] Confirm that the truck is in place and adjust the loading robot to perform loading according to the task execution plan;

[0012] Determine whether the loading robot has completed the loading work of the day, and obtain the actual loading volume data of the day to correct the LSTM model;

[0013] Obtain all loading information corresponding to the loading tasks of the day, and save the loading information as historical data in the database.

[0014] Optionally, the LSTM model includes an input layer, a hidden layer and an output layer;

[0015] The method of using the LSTM model to predict the loading volume data of the day according to the date data includes:

[0016] Using the input layer to read historical data of vehicle loading, preprocessing the historical data to obtain time series data;

[0017] The LSTM network of the hidden layer is used to learn the time series data, and the loading volume data of the day is predicted based on the historical data;

[0018] The output activation function of the output layer is used to output the predicted vehicle loading data.

[0019] Optionally, the process of modifying the LSTM model includes:

[0020] Compare the actual loading volume data with the predicted loading volume data, and obtain the loss value through the loss calculation function;

[0021] The loss value is fed back to the correction model of the hidden layer to perform error correction on the LSTM to reduce the error between the actual loading quantity data and the predicted loading quantity data.

[0022] Optionally, the environmental data includes weather conditions, temperature and humidity, wherein the weather conditions include sunny, rainy, snowy and foggy;

[0023] The loading task information includes the loading task type, cargo type, cargo quantity, single cargo weight, single cargo size, and cargo value;

[0024] The task types include loading from the warehouse floor to the transport vehicle compartment, and loading from the transport vehicle compartment to the warehouse floor;

[0025] The value of the goods includes expensive, ordinary and cheap;

[0026] The method of obtaining a task execution plan using a trained deep learning model according to the loading volume data, environmental data and loading task information includes:

[0027] The task type, cargo type, cargo quantity, single cargo weight, single cargo size, cargo value, predicted loading volume data and current weather conditions are input into the deep learning model to obtain the loading speed of the current task, the number of cargoes that the robot can grasp at a time and the unloading height of the robot.

[0028] Optionally, the task execution scheme further satisfies the following conditions:

[0029] If the temperature on that day is higher or lower than the storage temperature required for the goods, the loading speed is:

[0030]

[0031] If the humidity on that day is higher or lower than the storage humidity required for the goods, the loading speed is:

[0032]

[0033] Among them, T c is the temperature of the day, T is the storage temperature required for the goods, H c is the humidity of the day, H is the storage humidity required for the goods, V Deep is the loading speed of this task output by the deep learning model, V M is the maximum loading speed;

[0034] If the weather condition is snow, the loading speed shall not exceed half of the maximum loading speed;

[0035] If the temperature on that day is higher or lower than the storage temperature required for the goods, and the humidity on that day is higher or lower than the storage humidity required for the goods, and the weather condition is snowing, the minimum loading speed under each condition shall be taken as the loading speed.

[0036] Optionally, the task execution scheme further satisfies the following conditions:

[0037] When the robot grabs the cargo and reaches the unloading height, the robot is released to allow the pre-unloaded cargo to automatically fall to the ground or the carriage. The range of the unloading height is limited according to the value of the cargo:

[0038] When the cargo is valuable, the unloading height range is:

[0039] [L Min ,L Min +1 / 3*(LMax -L Min )]

[0040] When the cargo is of average value, the unloading height range is:

[0041] [L Min +1 / 3*(L Max -L Min ),L Min +2 / 3*(L Max -L Min )]

[0042] When the cargo is cheap, the unloading height range is:

[0043] [L Min +2 / 3*(L Max -L Min ),L Max ]

[0044] Among them, L Min is the minimum unloading height of the robot, L Max It is the maximum unloading height of the robot.

[0045] Optionally, the method further includes: when it is determined that the waiting time for the truck to arrive meets a judgment condition, notifying the administrator to manually formulate and execute a loading task plan, wherein the judgment condition is as follows:

[0046] The waiting time for the truck is greater than the first duration;

[0047] Or, the weather condition is sunny, and the temperature on that day is higher than the storage temperature, and the waiting time for the truck is greater than the second duration;

[0048] Or, the weather condition is rain or fog, and the humidity on that day is higher than the storage humidity, and the waiting time for the truck is greater than the second duration;

[0049] Or, the weather condition is snow, and the waiting time for the truck is greater than the second duration;

[0050] The first duration is greater than the second duration.

[0051] In a second aspect, the present application also provides a task management system for a loading robot, comprising:

[0052] The environment collection module is used to collect the current date data and the current environment data according to the loading task instructions;

[0053] An intelligent analysis module, for predicting the loading volume data of the day according to the date data using the LSTM model, and obtaining a task execution plan according to the loading volume data, environmental data and loading task information using a trained deep learning model;

[0054] The loading robot control module is used to determine whether the truck is in place and adjust the loading robot to perform loading according to the task execution plan;

[0055] The task management module is used to input, manage and extract loading task information, determine that the loading robot has completed the loading work of the day, and obtain the actual loading volume data of the day to correct the LSTM model;

[0056] The database module is used to obtain all loading information corresponding to the loading tasks of the day and save the loading information as historical data in the database.

[0057] In a third aspect, the present application provides an electronic device comprising: at least one memory for storing programs; and at least one processor for executing the programs stored in the memory. When the program stored in the memory is executed, the processor is used to execute the method described in the first aspect or any possible implementation of the first aspect.

[0058] In a fourth aspect, the present application provides a computer-readable storage medium, which stores a computer program. When the computer program runs on a processor, the processor executes the method described in the first aspect or any possible implementation of the first aspect.

[0059] In a fifth aspect, the present application provides a computer program product. When the computer program product runs on a processor, the processor executes the method described in the first aspect or any possible implementation of the first aspect.

[0060] It can be understood that the beneficial effects of the second to fifth aspects mentioned above can be found in the relevant description of the first aspect mentioned above, and will not be repeated here.

[0061] In general, the above technical solutions conceived by this application have the following beneficial effects compared with the prior art:

[0062] (1) This application introduces LSTM and deep learning models based on artificial intelligence technology. It aims to solve the problems of long loading and unloading time, easy damage of goods, unreasonable loading and unloading, and affected robot life. By accurately predicting the loading volume and analyzing environmental data in real time, it can optimize the loading speed and manipulator mode, thereby significantly reducing the loading and unloading time and the risk of damage to goods during the loading and unloading process. In addition, through intelligent task management and execution plan adjustment, it ensures the rationality of loading and unloading operations, improves the quality of loading and unloading, and thus extends the service life of the robot, which helps to improve the efficiency of the loading process and reduce the need for human intervention, thereby optimizing the overall operation process.

[0063] (2) This application takes into account multi-dimensional information such as environmental factors, task volume, and cargo characteristics, so that the loading robot can adjust its working mode in real time to adapt to complex and changing loading and unloading requirements. This not only improves the loading speed, but also effectively reduces the waiting time for cargo, ensuring the improvement of loading and unloading quality.

[0064] (3) This application saves loading information to the database, thus achieving effective management and utilization of historical data, which not only provides data support for subsequent task analysis and optimization, but also provides a reference for loading robots in future task execution. Through systematic data management, the service life of the robot can be further extended, maintenance costs can be reduced, and ultimately the intelligent and efficient loading and unloading operations can be achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 It is a flowchart of a task management method of a loading robot provided in an embodiment of the present application;

[0066] Figure 2 It is a mechanical layout structure diagram of the loading robot provided in an embodiment of the present application;

[0067] Figure 3 It is a structural diagram of the LSTM model of an embodiment of the present application;

[0068] Figure 4 is a structural diagram of a deep learning model in an embodiment of the present application;

[0069] Figure 5 It is a complete flow chart of the task management method of the loading robot provided in the embodiment of the present application;

[0070] Figure 6 It is a structural schematic diagram of a task management device for a loading robot provided in an embodiment of the present application;

[0071] Figure 7 It is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0072] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0073] The term "and / or" in this article is a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. The symbol " / " in this article indicates that the associated objects are in an or relationship, for example, A / B means A or B.

[0074] The terms "first" and "second" in the specification and claims herein are used to distinguish different objects rather than to describe a specific order of the objects. For example, a first response message and a second response message are used to distinguish different response messages rather than to describe a specific order of the response messages.

[0075] In the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0076] In the description of the embodiments of the present application, unless otherwise specified, "multiple" means two or more than two. For example, multiple processing units refer to two or more processing units, etc.; multiple elements refer to two or more elements, etc.

[0077] The embodiments of the present application are described below in conjunction with the drawings in the embodiments of the present application.

[0078] Reference Figure 1 The present application provides a task management method for a loading robot, comprising:

[0079] S101. Collect the current date data, and use the LSTM model to predict the loading volume data of the day based on the date data;

[0080] S102. Determine whether a user has issued a loading task instruction, collect current environmental data according to the loading task instruction, and extract loading task information;

[0081] S103. Obtain a task execution plan using a trained deep learning model based on loading volume data, environmental data, and loading task information;

[0082] S104. Determine that the truck is in place and adjust the loading robot to perform loading according to the task execution plan;

[0083] S105. Determine whether the loading robot has completed the loading work for the day, and obtain the actual loading volume data for the day to correct the LSTM model;

[0084] S106. Obtain all loading information corresponding to the loading tasks of the day, and save the loading information as historical data in the database.

[0085] Specifically, in S101, the current date data is first collected to lay the foundation for the subsequent loading volume forecast. By using the LSTM (Long Short-Term Memory) neural network model, it is possible to analyze historical loading volume data and identify potential time series patterns, thereby accurately predicting the loading volume for the day. The data-driven approach based on LSTM makes the forecast more accurate and can effectively respond to fluctuations in loading demand.

[0086] It should be noted that the advantage of the LSTM model lies in its ability to handle long-term dependencies, and it can capture the complex long-term relationship between loading volume and date. This process not only improves the accuracy of the prediction, but also provides an important basis for the subsequent loading task planning, ensuring that the system can respond at the appropriate time and improve overall efficiency.

[0087] Secondly, in S102, it is confirmed whether there is a loading task. This judgment basis ensures the rational allocation of resources and avoids unnecessary operations and waste. At the same time, the system will also collect current environmental data, such as temperature, humidity, and equipment status, in order to more comprehensively understand the execution conditions of the loading task.

[0088] In S103, the loading volume data, environmental data and loading task information are combined to generate a specific task execution plan using the trained deep learning model. This step takes into account multiple factors to ensure that the best loading strategy can be implemented in a complex environment.

[0089] Through the application of deep learning models, the task execution plan is not only based on historical data, but also can adapt to changes in the current environment. The dynamic adjustment capability improves the flexibility and adaptability of the loading robot, ensuring that the loading tasks can be efficiently performed under different circumstances and reducing the risk of human decision-making.

[0090] In S104, the process waits for the truck to arrive and then starts executing the task plan.

[0091] The judgment conditions for waiting for the truck to arrive are as follows:

[0092] 1. The waiting time for the truck is greater than the first duration;

[0093] 2. Meet any of the following conditions:

[0094] 1. The weather condition is sunny, the temperature on that day is higher than the storage temperature, and the waiting time for the truck is greater than the second duration;

[0095] 2. The weather condition is rain or fog, and the humidity on that day is higher than the storage humidity, and the waiting time for the truck is greater than the second duration;

[0096] 3. The weather condition is snow, and the waiting time for the truck is greater than the second duration;

[0097] The first duration is greater than the second duration.

[0098] The first duration is usually 10 minutes, and the second duration is 5 minutes. That is, if the waiting time for the truck exceeds ten minutes, or the above conditions exceed five minutes, the administrator will be notified for manual processing.

[0099] Furthermore, through S105, it is confirmed that the loading robot has completed the loading work for the day, and the actual loading volume data is obtained in time. This step is a feedback on the previous prediction and execution effect, ensuring the accuracy and completeness of the data.

[0100] The acquisition of actual loading volume data provides a basis for subsequent analysis and optimization. By comparing the predicted loading volume with the actual loading volume, the effectiveness of task execution can be evaluated and potential problems can be identified, thereby providing data support for subsequent decision-making.

[0101] It should be noted that, refer to Figure 2 The present application provides a loading robot, which adopts a truss structure and a manipulator to grab the cargo from the ground to the carriage of the transport vehicle, or from the carriage of the transport vehicle to the ground. The manipulator has the following features:

[0102] (1) The manipulator is mounted on a truss and suspended above the truck compartment;

[0103] (2) The manipulator relies on the truss to move forward and backward, left and right, and can move up and down, thus achieving three-degree-of-freedom movement;

[0104] (3) The manipulators are double-row and can pick up one bag of goods at a time or two bags of goods at a time.

[0105] Finally, S106 collects all relevant information about the day’s loading task and saves it to the database, which not only ensures the integrity and traceability of the data, but also provides valuable historical data for future analysis and model training.

[0106] Specifically, the saved loading information for the day includes the following: year, month, day, whether it is a holiday, whether it is a rest day, a list of loading task information for the day, the total number of goods loaded on the day, and the total weight of goods loaded on the day.

[0107] The list of loading task information for the day includes the task type, cargo type, cargo quantity, single cargo weight, single cargo size, cargo value, weather conditions, temperature, and humidity of each loading task on the day.

[0108] By saving loading information, we can better understand the changing trends and patterns in the loading and unloading process. The accumulation of these historical data will provide more training samples for subsequent deep learning models, thereby continuously improving the accuracy of predictions and the intelligence level of the system.

[0109] The embodiment of the present application introduces LSTM and deep learning models of artificial intelligence technology, and aims at the problems of long loading and unloading time, easy damage of goods, unreasonable loading and unloading, and affected robot life. By accurately predicting the loading volume and analyzing environmental data in real time, it can optimize the loading speed and manipulator mode, thereby significantly reducing the loading and unloading time and reducing the risk of damage to goods during the loading and unloading process. And through intelligent task management and execution plan adjustment, it ensures the rationality of loading and unloading operations, improves the loading and unloading quality, and thus extends the service life of the robot, which helps to improve the efficiency of the loading process and reduce the need for human intervention, thereby optimizing the overall operation process.

[0110] Reference Figure 3 , the LSTM model includes an input layer, a hidden layer and an output layer;

[0111] The method of using the LSTM model to predict the loading volume data of the day according to the date data includes:

[0112] Using the input layer to read historical data of vehicle loading, preprocessing the historical data to obtain time series data;

[0113] The LSTM network of the hidden layer is used to learn the time series data, and the loading volume data of the day is predicted based on the historical data;

[0114] The output activation function of the output layer is used to output the predicted vehicle loading data.

[0115] Also includes:

[0116] Compare the actual loading volume data with the predicted loading volume data, and obtain the loss value through the loss calculation function;

[0117] The loss value is fed back to the correction model of the hidden layer to perform error correction on the LSTM to reduce the error between the actual loading quantity data and the predicted loading quantity data.

[0118] Specifically, the process of prediction by the LSTM model in the embodiment of the present application is as follows:

[0119] First, the historical data of vehicle loading is read through the input layer. Historical data usually exists in the form of time series, which can reflect the trend of vehicle loading over time. By preprocessing these data, such as normalization, denoising, and filling missing values, they can be effectively converted into a format suitable for LSTM network processing. The purpose of preprocessing is to ensure the quality and consistency of the data, thereby improving the effect of subsequent model training.

[0120] The preprocessed time series data provides a clean input for the LSTM model, enabling it to capture the underlying patterns and trends in the data, ensuring that the model can be trained and predicted on valid data.

[0121] Secondly, the preprocessed time series data is input into the hidden layer LSTM network for training. Through its special structure, the LSTM network can effectively learn the long-term dependencies in the time series data and identify the changing patterns and periodic characteristics of the loading volume. After training, the LSTM network can generate a forecast of the loading volume for the day based on historical data. The LSTM forecast not only relies on the most recent loading volume data, but also takes into account earlier historical data, thereby improving the accuracy and reliability of the forecast.

[0122] After the model training is completed, the output layer of the LSTM network is responsible for processing the prediction results and outputting the predicted loading data. The output layer usually uses an appropriate activation function (such as a linear activation function) to ensure that the output value is consistent with the range of actual loading.

[0123] Furthermore, after obtaining the predicted loading volume, it is compared with the actual loading volume of the day. Through the loss calculation function, the gap between the predicted value and the actual value can be quantified to obtain the loss value. The loss value is an important indicator for evaluating model performance and can reflect the effectiveness of the model in practical applications.

[0124] Through comparative analysis, the performance of the model in specific situations can be identified, thus providing a basis for subsequent adjustments and optimizations.

[0125] Finally, the calculated loss value is fed back to the correction model of the hidden layer for error correction. The feedback mechanism enables the model to identify and adjust the systematic deviations in the prediction, thereby reducing the error between the predicted loading volume and the actual loading volume. Through the continuous iterative process, the model can gradually learn and optimize its prediction ability.

[0126] The error correction mechanism of this embodiment not only improves the adaptability and accuracy of the model, but also provides a more solid foundation for future prediction tasks. With the continuous accumulation of historical data and continuous optimization of the model, more efficient and accurate loading volume prediction can be achieved.

[0127] Optionally, the environmental data includes weather conditions, temperature and humidity, wherein the weather conditions include sunny, rainy, snowy and foggy;

[0128] The loading task information includes the loading task type, cargo type, cargo quantity, single cargo weight, single cargo size, and cargo value;

[0129] The task types include loading from the warehouse floor to the transport vehicle compartment, and loading from the transport vehicle compartment to the warehouse floor;

[0130] The value of the goods includes expensive, ordinary and cheap;

[0131] Reference Figure 4 , the step of obtaining a task execution plan using a trained deep learning model according to the loading volume data, environmental data and loading task information includes:

[0132] The task type, cargo type, cargo quantity, single cargo weight, single cargo size, cargo value, predicted loading volume data and current weather conditions are input into the deep learning model to obtain the loading speed of the current task, the number of cargoes that the robot can grasp at a time and the unloading height of the robot.

[0133] Specifically, the environmental data in the embodiments of the present application is an important factor affecting loading efficiency and cargo safety. Specifically, weather conditions, such as sunny, rainy, snowy or foggy, directly affect loading and unloading work, especially in the case of outdoor operations.

[0134] For example, rainy days may cause the goods to get damp during loading and unloading, snowy days may make the ground slippery, increasing the risk of operation; foggy days may affect the visual distance and safety, and additional precautions need to be taken. In addition, temperature and humidity can also affect the storage conditions of certain goods. Sparse standard goods such as food or medicine are highly dependent on environmental control.

[0135] Loading task information involves multiple dimensions, including task type, cargo type, cargo quantity, single cargo weight, single cargo size, and cargo value.

[0136] Furthermore, the weight, size and value of the cargo are equally important. Heavier or bulkier cargo may require more powerful manipulators and slower loading speeds, while valuable cargo may need to be handled with extra care to avoid damage or loss.

[0137] After obtaining environmental data and loading task information, the trained deep learning model uses this information as input to generate the corresponding task execution plan. Specifically, the model uses task type, cargo type, cargo quantity, single cargo weight, single cargo size, cargo preciousness, predicted loading volume, and current weather conditions as input features, and learns and infers through a multi-layer network structure.

[0138] The deep learning model can provide specific execution parameters, including loading speed, the number of goods the robot can pick up at a time, and the robot's unloading height, which not only improves the efficiency of loading operations but also ensures the safety of cargo handling.

[0139] The embodiment of the present application can generate an efficient loading execution plan through a deep learning model, thereby improving the overall loading efficiency and safety. It not only optimizes resource allocation, but also can reduce manual intervention and improve work efficiency.

[0140] Optionally, the task execution plan also satisfies the following conditions:

[0141] If the temperature on that day is higher or lower than the storage temperature required for the goods, the loading speed is:

[0142]

[0143] If the humidity on that day is higher or lower than the storage humidity required for the goods, the loading speed is:

[0144]

[0145] Among them, T c is the temperature of the day, T is the storage temperature required for the goods, H c is the humidity of the day, H is the storage humidity required for the goods, V Deep is the loading speed of this task output by the deep learning model, V M is the maximum loading speed;

[0146] If the weather condition is snow, the loading speed shall not exceed half of the maximum loading speed;

[0147] If the temperature on that day is higher or lower than the storage temperature required for the goods, and the humidity on that day is higher or lower than the storage humidity required for the goods, and the weather condition is snowing, the minimum loading speed under each condition shall be taken as the loading speed.

[0148] Optionally, it also includes:

[0149] When the robot grabs the cargo and reaches the unloading height, the robot is released to allow the pre-unloaded cargo to automatically fall to the ground or the carriage. The range of the unloading height is limited according to the value of the cargo:

[0150] When the cargo is valuable, the unloading height range is:

[0151] [L Min ,L Min +1 / 3*(L Max -L Min )]

[0152] When the cargo is of average value, the unloading height range is:

[0153] [L Min +1 / 3*(L Max -L Min ),L Min +2 / 3*(L Max -L Min )]

[0154] When the cargo is cheap, the unloading height range is:

[0155] [L Min +2 / 3*(L Max -L Min ),L Max ]

[0156] Among them, L Min is the minimum unloading height of the robot, L Max It is the maximum unloading height of the robot.

[0157] Specifically, in order to fully ensure the safety of cargo loading and unloading, the maximum loading speed of the robot is set to V M , the loading speed of this task output by the deep learning model is V Deep , then the loading execution plan formulated shall refer to Figure 4 The output of a deep learning model must also comply with the following rules:

[0158] (1) If the temperature on that day is T c If the temperature is higher or lower than the storage temperature T required for the goods, the loading speed C is:

[0159]

[0160] (2) If the humidity on that day is H c If the humidity is higher or lower than the storage humidity H required by the goods, the loading speed C is:

[0161]

[0162] (3) If it is snowing, to prevent ice and slipping, the loading speed C should not be greater than V M 50% of.

[0163] (4) When multiple rules (1), (2) and (3) occur simultaneously, the minimum value of C calculated by multiple rules is taken as the loading speed of this task.

[0164] (5) When the robot unloads the cargo grabbed in the air to the ground or the carriage, it does not need to descend to the same height as the ground or the carriage to release the cargo. Instead, it can release the cargo at a certain height and let it fall to the ground or the carriage below. This can reduce the time of cargo unloading. The higher the height of the robot unloading, the shorter the cargo unloading time and the less time it takes to load the cargo. However, the robot may damage the cargo if it drops the cargo from a very high position, so the robot unloading height needs to be adjusted according to the value of the cargo, that is:

[0165] When the cargo is valuable, the unloading height range is:

[0166] [L Min ,L Min +1 / 3*(L Max -L Min )]

[0167] When the cargo is of average value, the unloading height range is:

[0168] [L Min +1 / 3*(L Max -L Min ),L Min +2 / 3*(L Max -L Min )]

[0169] When the cargo is cheap, the unloading height range is:

[0170] [L Min +2 / 3*(L Max -L Min ),L Max ]

[0171] Among them, L Min is the minimum unloading height of the robot, L Max It is the maximum unloading height of the robot.

[0172] Reference Figure 5 , Figure 5 It is a complete flow chart of the embodiment of the present application, which includes the following steps:

[0173] S1. Collect current date data

[0174] S2, use the LSTM model to predict the loading volume of the day;

[0175] S3, determine whether there is a loading task at present, if yes, go to step S4, otherwise go to step S3;

[0176] S4, collect current environmental data;

[0177] S5, extracting loading task information;

[0178] S6. Use deep learning models to analyze the execution plan of the current task based on environmental data and loading task information;

[0179] S7, waiting for the truck to arrive, if it arrives, go to step S8, otherwise go to step S7;

[0180] S8, adjusting the loading robot according to the task execution plan to execute loading;

[0181] S9, determine whether the robot has finished working for the day, if yes, go to step S10, if not, go to step S3;

[0182] S10, using the actual loading volume of the day to correct the LSTM model and update the LSTM prediction model;

[0183] S11, saving the loading information of the day as historical data in the database;

[0184] S12: The robot finishes its work and shuts down.

[0185] Reference Figure 6 The present application also provides a task management system for a loading robot, including:

[0186] The environment collection module is used to collect the current date data and the current environment data according to the loading task instructions;

[0187] An intelligent analysis module, for predicting the loading volume data of the day according to the date data using the LSTM model, and obtaining a task execution plan according to the loading volume data, environmental data and loading task information using a trained deep learning model;

[0188] The loading robot control module is used to determine whether the truck is in place and adjust the loading robot to perform loading according to the task execution plan;

[0189] The task management module is used to input, manage and extract loading task information, determine that the loading robot has completed the loading work of the day, and obtain the actual loading volume data of the day to correct the LSTM model;

[0190] The database module is used to obtain all loading information corresponding to the loading tasks of the day and save the loading information as historical data in the database.

[0191] Optionally, the LSTM model includes an input layer, a hidden layer and an output layer;

[0192] The method of using the LSTM model to predict the loading volume data of the day according to the date data includes:

[0193] Using the input layer to read historical data of vehicle loading, preprocessing the historical data to obtain time series data;

[0194] The LSTM network of the hidden layer is used to learn the time series data, and the loading volume data of the day is predicted based on the historical data;

[0195] The output activation function of the output layer is used to output the predicted vehicle loading data.

[0196] Optionally, the system further comprises a comparison module, configured to:

[0197] Compare the actual loading volume data with the predicted loading volume data, and obtain the loss value through the loss calculation function;

[0198] The loss value is fed back to the correction model of the hidden layer for error correction to reduce the error between the actual loading quantity data and the predicted loading quantity data.

[0199] Optionally, the environmental data includes weather conditions, temperature and humidity, wherein the weather conditions include sunny, rainy, snowy and foggy;

[0200] The loading task information includes the loading task type, cargo type, cargo quantity, single cargo weight, single cargo size, and cargo value;

[0201] The task types include loading from the warehouse floor to the transport vehicle compartment, and loading from the transport vehicle compartment to the warehouse floor;

[0202] The value of the goods includes expensive, ordinary and cheap;

[0203] The method of obtaining a task execution plan using a trained deep learning model according to the loading volume data, environmental data and loading task information includes:

[0204] The task type, cargo type, cargo quantity, single cargo weight, single cargo size, cargo value, predicted loading volume data and current weather conditions are input into the deep learning model to obtain the loading speed of the current task, the number of cargoes that the robot can grasp at a time and the unloading height of the robot.

[0205] Optionally, the task execution plan also satisfies the following conditions:

[0206] If the temperature on that day is higher or lower than the storage temperature required for the goods, the loading speed is:

[0207]

[0208] If the humidity on that day is higher or lower than the storage humidity required for the goods, the loading speed is:

[0209]

[0210] Among them, T c is the temperature of the day, T is the storage temperature required for the goods, H c is the humidity of the day, H is the storage humidity required for the goods, V Deep is the loading speed of this task output by the deep learning model, V M is the maximum loading speed;

[0211] If the weather condition is snow, the loading speed shall not exceed half of the maximum loading speed;

[0212] If the temperature on that day is higher or lower than the storage temperature required for the goods, and the humidity on that day is higher or lower than the storage humidity required for the goods, and the weather condition is snowing, the minimum loading speed under each condition shall be taken as the loading speed.

[0213] Optionally, an uninstall control is also included for:

[0214] When the robot grabs the cargo and reaches the unloading height, the robot is released to allow the pre-unloaded cargo to automatically fall to the ground or the carriage. The range of the unloading height is limited according to the value of the cargo:

[0215] When the cargo is valuable, the unloading height range is:

[0216] [L Min ,L Min +1 / 3*(L Max -L Min )]

[0217] When the cargo is of average value, the unloading height range is:

[0218] [L Min +1 / 3*(L Max -L Min ),L Min +2 / 3*(L Max -L Min )]

[0219] When the cargo is cheap, the unloading height range is:

[0220] [L Min +2 / 3*(L Max -L Min ),LMax ]

[0221] Among them, L Min is the minimum unloading height of the robot, L Max It is the maximum unloading height of the robot.

[0222] Optionally, the method further includes: when it is determined that the waiting time for the truck to arrive meets a judgment condition, notifying the administrator to manually formulate and execute a loading task plan, wherein the judgment condition is as follows:

[0223] The waiting time for the truck is greater than the first duration;

[0224] Or, the weather condition is sunny, and the temperature on that day is higher than the storage temperature, and the waiting time for the truck is greater than the second duration;

[0225] Or, the weather condition is rain or fog, and the humidity on that day is higher than the storage humidity, and the waiting time for the truck is greater than the second duration;

[0226] Or, the weather condition is snow, and the waiting time for the truck is greater than the second duration;

[0227] The first duration is greater than the second duration.

[0228] It should be understood that the above-mentioned device is used to execute the method in the above-mentioned embodiment. The implementation principle and technical effect of the corresponding program module in the device are similar to those described in the above-mentioned method. The working process of the device can refer to the corresponding process in the above-mentioned method, which will not be repeated here.

[0229] Reference Figure 7 Based on the method in the above embodiment, the embodiment of the present application provides an electronic device, which may include: a processor (Processor) 710, a communication interface (CommunicationsInterface) 720, a memory (Memory) 730 and a communication bus 740, wherein the processor 710, the communication interface 720, and the memory 730 communicate with each other through the communication bus 740. The processor 710 can call the logic instructions in the memory 730 to execute the method in the above embodiment.

[0230] In addition, the logic instructions in the above-mentioned memory 730 can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application.

[0231] Based on the method in the above embodiment, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program. When the computer program runs on a processor, the processor executes the method in the above embodiment.

[0232] Based on the method in the above embodiment, an embodiment of the present application provides a computer program product. When the computer program product runs on a processor, the processor executes the method in the above embodiment.

[0233] It is understandable that the processor in the embodiment of the present application may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. The general-purpose processor may be a microprocessor or any conventional processor.

[0234] The method steps in the embodiments of the present application can be implemented by hardware or by a processor executing software instructions. The software instructions can be composed of corresponding software modules, and the software modules can be stored in random access memory (RAM), flash memory, read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), registers, hard disks, mobile hard disks, CD-ROMs, or any other form of storage medium known in the art. An exemplary storage medium is coupled to a processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can be located in an ASIC.

[0235] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted through the computer-readable storage medium. The computer instructions may be transmitted from a website site, a computer, a server or a data center to another website site, a computer, a server or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or a data center that includes one or more available media integrated. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)), etc.

[0236] It should be understood that the various numerical numbers involved in the embodiments of the present application are only used for the convenience of description and are not used to limit the scope of the embodiments of the present application.

[0237] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A task management method for a loading robot, characterized in that: include: Collect current date data, and use the LSTM model to predict the loading volume data of the day based on the date data; Determine whether a user has issued a loading task instruction, collect current environmental data according to the loading task instruction, and extract loading task information; Obtain the task execution plan using the trained deep learning model based on the loading volume data, environmental data, and loading task information; Confirm that the truck is in place and adjust the loading robot to perform loading according to the task execution plan; Determine whether the loading robot has completed the loading work of the day, and obtain the actual loading volume data of the day to correct the LSTM model; Obtain all loading information corresponding to the loading tasks of the day, and save the loading information as historical data in the database.

2. The task management method of the loading robot according to claim 1, characterized in that: The LSTM model includes an input layer, a hidden layer and an output layer; The method of using the LSTM model to predict the loading volume data of the day according to the date data includes: Using the input layer to read historical data of vehicle loading, preprocessing the historical data to obtain time series data; The LSTM network of the hidden layer is used to learn the time series data, and the loading volume data of the day is predicted based on the historical data; The output activation function of the output layer is used to output the predicted vehicle loading data.

3. The task management method of the loading robot according to claim 2, characterized in that: The process of modifying the LSTM model includes: Compare the actual loading volume data with the predicted loading volume data, and obtain the loss value through the loss calculation function; The loss value is fed back to the correction model of the hidden layer to perform error correction on the LSTM to reduce the error between the actual loading quantity data and the predicted loading quantity data.

4. The task management method of the loading robot according to claim 1, characterized in that: The environmental data includes weather conditions, temperature and humidity, wherein the weather conditions include sunny, rainy, snowy and foggy; The loading task information includes the loading task type, cargo type, cargo quantity, single cargo weight, single cargo size, and cargo value; The task types include loading from the warehouse floor to the transport vehicle compartment, and loading from the transport vehicle compartment to the warehouse floor; The value of the goods includes expensive, ordinary and cheap; The method of obtaining a task execution plan using a trained deep learning model according to the loading volume data, environmental data and loading task information includes: The task type, cargo type, cargo quantity, single cargo weight, single cargo size, cargo value, predicted loading volume data and current weather conditions are input into the deep learning model to obtain the loading speed of the current task, the number of cargoes that the robot can grasp at a time and the unloading height of the robot.

5. The task management method of the loading robot according to claim 4, characterized in that: The task execution scheme also meets the following conditions: If the temperature on that day is higher or lower than the storage temperature required for the goods, the loading speed is: If the humidity on that day is higher or lower than the storage humidity required for the goods, the loading speed is: Among them, T c is the temperature of the day, T is the storage temperature required for the goods, H c is the humidity of the day, H is the storage humidity required for the goods, V Deep is the loading speed of this task output by the deep learning model, V M is the maximum loading speed; If the weather condition is snow, the loading speed shall not exceed half of the maximum loading speed; If the temperature on that day is higher or lower than the storage temperature required for the goods, and the humidity on that day is higher or lower than the storage humidity required for the goods, and the weather condition is snowing, the minimum loading speed under each condition shall be taken as the loading speed.

6. The task management method of the loading robot according to claim 4, characterized in that: The task execution scheme also meets the following conditions: When the robot grabs the cargo and reaches the unloading height, the robot is released to allow the pre-unloaded cargo to automatically fall to the ground or the carriage. The range of the unloading height is limited according to the value of the cargo: When the cargo is valuable, the unloading height range is: [L Min ,L Min +1 / 3*(L Max -L Min )] When the cargo is of average value, the unloading height range is: [L Min +1 / 3*(L Max -L Min ),L Min +2 / 3*(L Max -L Min )] When the cargo is cheap, the unloading height range is: [L Min +2 / 3*(L Max -L Min ),L Max ] Among them, L Min is the minimum unloading height of the robot, L Max It is the maximum unloading height of the robot.

7. The task management method of the loading robot according to claim 4, characterized in that: Also includes: When it is determined that the waiting time for the truck to arrive meets the judgment conditions, the administrator is notified to manually formulate and execute the loading task plan. The judgment conditions are as follows: The waiting time for the truck is greater than the first duration; Or, the weather condition is sunny, and the temperature on that day is higher than the storage temperature, and the waiting time for the truck is greater than the second duration; Or, the weather condition is rain or fog, and the humidity on that day is higher than the storage humidity, and the waiting time for the truck is greater than the second duration; Or, the weather condition is snow, and the waiting time for the truck is greater than the second duration; The first duration is greater than the second duration.

8. A task management system for a loading robot, characterized in that: include: The environment collection module is used to collect the current date data and the current environment data according to the loading task instructions; An intelligent analysis module, for predicting the loading volume data of the day according to the date data using the LSTM model, and obtaining a task execution plan according to the loading volume data, environmental data and loading task information using a trained deep learning model; The loading robot control module is used to determine whether the truck is in place and adjust the loading robot to perform loading according to the task execution plan; The task management module is used to input, manage and extract loading task information, determine that the loading robot has completed the loading work of the day, and obtain the actual loading volume data of the day to correct the LSTM model; The database module is used to obtain all loading information corresponding to the loading tasks of the day and save the loading information as historical data in the database.

9. An electronic device, characterized in that: include: at least one memory for storing a computer program; At least one processor is used to execute the program stored in the memory. When the program stored in the memory is executed, the processor is used to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program runs on a processor, the processor is caused to execute the method according to any one of claims 1 to 7.

Images