Grain storage and transportation disorder plan autonomous scheduling system and method

Abstract

The invention relates to the field of intelligent optimization systems and application thereof, in particular to a grain storage and transportation disorder plan autonomous scheduling system and method. Comprising the following steps: carrying out calibration and time synchronization on collected data, and carrying out data cleaning, screening, redundancy removal and anomaly detection processing; further performing encryption and structuring processing; constructing a prediction model based on the encrypted structured data, generating prediction output, dynamically optimizing and updating model parameters, and evaluating the prediction performance of the prediction model; an optimization task scheduling scheme is output based on the data after anomaly detection and prediction; calculating performance evaluation indexes and generating optimization suggestions; a new resource allocation strategy and a task scheduling scheme are provided, and functions of monitoring system performance in real time, adjusting an interface of system parameters and the like are provided for users. The problem that the accuracy and the adaptability are low when dynamic data are processed is solved. And in the face of complex and dynamic tasks, resource allocation cannot be adjusted in real time to adapt to constantly changing requirements.

Description

Technical Field

[0001] The present invention relates to an intelligent optimization system and its application field, and in particular to an autonomous dispatching system and method for disorderly planning of grain storage and transportation. Background Art

[0002] In traditional grain storage and transportation management, the approach that relies on fixed scheduling rules and empirical judgment has shown its limitations. This approach cannot flexibly respond to dynamically changing production and transportation needs, resulting in limited effectiveness and real-time planning. At the same time, in traditional grain storage and transportation management, resource scheduling and allocation also have problems of inefficiency and inconsistency, and often fail to meet the complex needs of actual operations. In addition, the grain storage and transportation process involves the processing and analysis of a large amount of dynamic information, and traditional manual operations or calculation methods are difficult to achieve efficient and accurate scheduling decisions. In order to improve the efficiency of grain storage management and optimize resource allocation, it is urgent to develop an autonomous scheduling system and method for grain storage and transportation based on intelligent technology to achieve more efficient operations and better resource utilization. Summary of the invention

[0003] The present invention provides an autonomous scheduling system and method for disordered storage and transportation of grain silos to solve the problems of low accuracy and adaptability when processing dynamic data; and the inability to adjust resource allocation in real time to adapt to changing needs when facing complex and dynamic tasks, resulting in low resource utilization efficiency and non-optimized scheduling.

[0004] The present invention provides a system and method for autonomously dispatching disordered grain storage and transportation plans, which specifically includes the following technical solutions: A method for autonomously scheduling disordered grain storage and transportation plans, comprising the following steps: S1. Collect data in real time, calibrate and time synchronize the data, clean, filter, remove redundancy and detect anomalies on the synchronized data; encrypt and structure the data after anomaly detection to obtain encrypted structured data, and store the encrypted structured data in the database; S2. Build a prediction model based on the encrypted structured data, generate prediction output, and use an adaptive incremental optimization algorithm to dynamically optimize and update the model parameters of the prediction model; use an improved relative error evaluation index to evaluate the prediction performance of the prediction model, and store the updated model parameters and prediction results in the database; S3. Based on the data and prediction output after anomaly detection, the task scheduling scheme is optimized through a dynamic adaptive fusion optimization algorithm; S4. Calculate performance evaluation indicators to evaluate the performance of the scheduling system in resource scheduling and task execution; generate optimization suggestions based on performance evaluation indicators; propose new resource allocation strategies and task scheduling solutions, and provide users with an interface for real-time monitoring of system performance and adjustment of system parameters, feedback prompts and suggestions, and historical data query functions.

[0005] Preferably, the S1 specifically includes: The collected data is transmitted to the central server or data receiving terminal, and the sensor output is initially calibrated using linear regression; the time synchronization algorithm is applied to adjust the timestamp of the data to match the server time, and the synchronized data is cleaned using the moving average algorithm to obtain smoothed data; the smoothed data is then filtered according to the set threshold range; redundancy is removed based on the filtered data, and duplicate data is identified and deleted; abnormal data in the data after redundancy is removed is identified through Z-score anomaly detection.

[0006] Preferably, the S2 specifically includes: Decrypt the encrypted structured data to obtain the decrypted data, and convert the decrypted data into a time series data matrix; based on the decrypted data, reconstruct the phase space data through the time delay embedding method; then use the nonlinear phase space mapping function Construct a prediction model. The specific formula is: , in, is the predicted output, indicating that The model predicted value of is the timestamp after synchronization; Represents phase space data; represents the nonlinear phase space mapping function; is the weight matrix; are the parameters in the weight matrix; is a bias term used to adjust the predicted output; Represents the activation function.

[0007] Preferably, the S2 specifically includes: The specific formula of the adaptive incremental optimization algorithm is: , in, are the model parameters of the updated prediction model; It is Actual values, representing the true target variable value; is the square norm of the weight matrix and the parameters in the weight matrix, used for regularization; For predictions using phase space data; Indicates time point; is the regularization coefficient; is the number of actual values; is the actual value index.

[0008] Preferably, the S3 specifically includes: In the implementation process of the dynamic adaptive fusion optimization algorithm, data fusion processing is performed based on the abnormal detection data obtained from S1 and the prediction output obtained from S2 to obtain the fused data; and the scheduling objective function is defined by weighted summation based on the cost, completion time, resource utilization and fused data. The formula is as follows: , in, is the scheduling objective function; is the resource allocation variable; is the task scheduling variable; is the fused data; is the cost function; is the completion time function; is a resource utilization function; , , It is the weight coefficient of different optimization objectives in the scheduling objective function.

[0009] Preferably, the S3 specifically includes: In the implementation process of the dynamic adaptive fusion optimization algorithm, a dynamic adjustment mechanism is introduced to adjust the weight coefficient. The specific formula is: , in, is the dynamically adjusted weight coefficient; is the initial value of the weight coefficient; It is a parameter used to dynamically adjust the weights and is used to control the impact of real-time feedback errors on weight adjustment. is the real-time feedback error; Indicates Parameters related to the sensitivity of the allocation amount of each resource; It is a constant that prevents infinity or negative numbers from appearing in calculations.

[0010] Preferably, the S3 specifically includes: In the implementation process of the dynamic adaptive fusion optimization algorithm, the task scheduling scheme is optimized by minimizing the scheduling objective function and generating the optimized scheduling objective function; the scheduling system performance is evaluated by calculating the expected value of the scheduling objective function predetermined by the scheduling system and the error of the optimized scheduling objective function.

[0011] Preferably, the S4 specifically includes: By analyzing the errors, the performance evaluation indicators are calculated and a system performance evaluation report is generated, including the average error, maximum error and minimum error.

[0012] Preferably, the S4 specifically includes: Generate a report based on the error analysis results and optimization strategy for reference by the scheduling system administrator or user; the optimization strategy is a strategy formulated based on the error analysis, including a new resource allocation strategy, adjusting the learning rate, and adjusting the regularization parameter; the feedback data and performance evaluation indicators are then persistently stored, and the feedback data is encrypted; the encrypted feedback data is stored in the database and cleaned up regularly.

[0013] An autonomous dispatching system for disorderly storage and transportation of grain silos, comprising the following parts: Sensor network module, data acquisition and processing module, historical data analysis module, intelligent scheduling engine module, execution control module, feedback and optimization module, human-computer interaction interface module; The sensor network module is used to collect data on the environment, grain quality, equipment status and location in the warehouse and during transportation, perform automatic calibration and self-diagnosis in real time, add timestamps to all collected data for synchronization, and provide the synchronized data to the data collection and processing module; The data acquisition and processing module is used to receive the synchronized data from the sensor network module, and perform cleaning, screening, redundancy removal and anomaly detection processing. The data after anomaly detection is structured and stored in the database and provided to the historical data analysis module and the intelligent scheduling engine module; The historical data analysis module builds a prediction model based on the structured data after anomaly detection stored in the data acquisition and processing module, and dynamically updates and optimizes the prediction model using the historical data accumulated by the system over a long period of time, and provides the prediction results to the intelligent scheduling engine module; The intelligent scheduling engine module is used to receive the data after abnormal detection of the data acquisition and processing module, the prediction results of the historical data analysis module, the feedback information of the execution control module, the optimization suggestions and feedback data of the feedback and optimization module, and the adjustment instructions input by the user of the human-computer interaction interface module, and optimize the task scheduling plan by using the dynamic adaptive fusion optimization algorithm; and provide the optimized scheduling plan to the execution control module and the human-computer interaction interface module; The execution control module is used to receive the optimized scheduling plan of the intelligent scheduling engine module and the adjustment instructions input by the user of the human-computer interaction interface module, and convert them into execution commands and send them to the execution unit, monitor the execution of the commands in real time and identify abnormal situations, automatically adjust the execution commands, and provide the recorded command execution and abnormal report feedback information to the intelligent scheduling engine module, the feedback and optimization module and the human-computer interaction interface module; The feedback and optimization module is used to receive feedback information from the execution control module, compare and evaluate the expected and actual execution results, analyze the feedback information to generate optimization suggestions and store the feedback data persistently, and provide the optimization suggestions and feedback data to the intelligent scheduling engine module; The human-computer interaction interface module is used to receive the optimized scheduling plan of the intelligent scheduling engine module, the feedback information of the execution control module and the adjustment instructions input by the user, and provide a visual interface to display the system operation status, scheduling plan and execution status, allowing the user to monitor the system in real time, adjust parameters or perform emergency intervention, and provide the adjustment instructions input by the user to the intelligent scheduling engine module and the execution control module.

[0014] The beneficial effects of the technical solution of the present invention are: 1. Real-time data collection through high-precision sensors, combined with automatic calibration, time synchronization, data cleaning, redundancy removal and exception handling, ensures data accuracy and consistency. At the same time, encrypted storage is used to improve data security and privacy protection. These measures jointly improve the reliability and processing efficiency of data in the scheduling system in complex environments, and help optimize the overall performance of resource scheduling and task execution.

[0015] 2. The prediction model is constructed through phase space reconstruction and nonlinear phase space mapping function, and the model parameters are dynamically updated through the adaptive incremental optimization algorithm to improve the accuracy of the prediction. Combined with the improved relative error evaluation index, the model performance is accurately evaluated to achieve accurate prediction and optimization of complex time series data, providing a reliable basis for real-time monitoring and adjustment, and enhancing the prediction ability and robustness of the scheduling system.

[0016] 3. Through the dynamic adaptive fusion optimization algorithm, the task scheduling scheme is dynamically optimized, which enhances the adaptability and efficiency of the scheduling system in complex environments, ensures the optimal allocation of resources and efficient execution of tasks, and improves the overall system performance and flexibility.

[0017] 4. Through performance evaluation and optimization suggestions, real-time monitoring and adjustment of resource scheduling and task execution of the scheduling system can be carried out to improve the accuracy and efficiency of the scheduling system. By encrypting and storing feedback data and providing a multi-functional user interface, it supports real-time monitoring, parameter adjustment and historical data query to ensure the efficiency and reliability of the scheduling system in actual applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a module composition diagram of a grain storage and transportation disorderly plan autonomous scheduling system according to the present invention; Figure 2 The present invention provides a flow chart of an autonomous dispatching method for disorderly planned grain storage and transportation. DETAILED DESCRIPTION

[0019] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the technical scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is only a part of the embodiment of the present invention, not all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0021] The specific scheme of the autonomous dispatching system and method for disorderly planning of grain storage and transportation provided by the present invention is described in detail below with reference to the accompanying drawings.

[0022] See attached Figure 1 , which shows a module composition diagram of a grain storage and transportation disorderly plan autonomous scheduling system provided by an embodiment of the present invention, including the following parts: Sensor network module, data acquisition and processing module, historical data analysis module, intelligent scheduling engine module, execution control module, feedback and optimization module, human-computer interaction interface module; The sensor network module is used to collect the environment, grain quality, equipment status and location data in the warehouse and during transportation through various sensors, perform automatic calibration and self-diagnosis in real time, add timestamps to all collected data for synchronization, and provide the synchronized data to the data collection and processing module; The data acquisition and processing module is used to receive the synchronized data from the sensor network module, and perform cleaning, screening, redundancy removal and anomaly detection processing. The data after anomaly detection is structured and stored in the database and provided to the historical data analysis module and the intelligent scheduling engine module; The historical data analysis module constructs a prediction model based on the structured stored data after abnormal detection in the data acquisition and processing module, and dynamically updates and optimizes the prediction model by using the historical data accumulated by the system for a long time, and provides the prediction results to the intelligent scheduling engine module; The intelligent scheduling engine module is used to receive the data after abnormal detection of the data acquisition and processing module, the prediction results of the historical data analysis module, the feedback information of the execution control module, the optimization suggestions and feedback data of the feedback and optimization module, and the adjustment instructions input by the user of the human-computer interaction interface module, and optimize the task scheduling plan by using the dynamic adaptive fusion optimization algorithm; and provide the optimized scheduling plan to the execution control module and the human-computer interaction interface module; The execution control module is used to receive the optimized scheduling scheme of the intelligent scheduling engine module and the adjustment instructions input by the user of the human-computer interaction interface module, and convert them into specific execution commands and send them to the relevant execution units, monitor the execution of commands in real time and identify abnormal situations, automatically adjust the execution commands, and provide the recorded command execution and abnormal report feedback information to the intelligent scheduling engine module, the feedback and optimization module and the human-computer interaction interface module; The feedback and optimization module is used to receive feedback information from the execution control module, compare and evaluate expected and actual execution results, analyze the feedback information to generate optimization suggestions and store the feedback data persistently, and provide the optimization suggestions and feedback data to the intelligent scheduling engine module; The human-computer interaction interface module is used to receive the optimized scheduling plan of the intelligent scheduling engine module, the feedback information of the execution control module and the adjustment instructions input by the user, and provide an intuitive visual interface to display the system operation status, scheduling plan and execution status, allowing the user to monitor the system in real time, adjust parameters or perform emergency intervention, and provide the adjustment instructions input by the user to the intelligent scheduling engine module and the execution control module.

[0023] See attached Figure 2 , which shows a flow chart of a method for autonomously scheduling grain storage and transportation disorderly plans provided by an embodiment of the present invention, the method comprising the following steps: S1. Collect data in real time, calibrate and time synchronize the data, clean, filter, remove redundancy and detect anomalies on the synchronized data; encrypt and structure the data after anomaly detection to obtain encrypted structured data, and store the encrypted structured data in the database; Select suitable high-precision sensors, such as temperature and humidity sensors, pressure sensors, position sensors, etc., and install them in the specified locations. Set the acquisition frequency to The sampling time interval is , data collection is performed based on sampling time intervals, and the collected data is transmitted to a central server or data receiving terminal via a wireless or wired network.

[0024] , in, For sensor In time Data collected; Indicates the sensor at time The actual measured value of the collected data.

[0025] An initial calibration of the sensor output was performed using linear regression to adjust for measurement bias.

[0026] , in, It's in time The calibrated data; is the calibration factor; is the calibration offset.

[0027] Apply the time synchronization algorithm to adjust the timestamp of the data to match the server time, ensure that the data collected by the sensor is consistent with the server time, and add metadata such as timestamp and sensor ID to the data.

[0028] , in, is the timestamp after synchronization; is the server timestamp; is the sensor timestamp; It is the local time.

[0029] The synchronized data is cleaned using a moving average algorithm to detect and remove invalid data, such as sensor failure data or lost data.

[0030] , in, is the smoothed data; is the calibrated data; is the size of the smoothing window, that is, the number of samples. Samples refer to the calibrated data points involved in the calculation within the smoothing window, which come from the collected data; is the index within the window; is the time step in the window; Indicates the calibrated data at time Then, according to the step length Offset The corresponding data at each time point.

[0031] Then, the data is screened according to the threshold range set based on expert experience to filter out the data that exceeds the range.

[0032] , in, It is the filtered data; is the lower filtering threshold, is the upper filter threshold.

[0033] Then perform redundancy removal to identify and delete duplicate data.

[0034] , in, Function is used to remove duplicate data; To remove redundant data.

[0035] Z-score anomaly detection identifies abnormal data in the data after redundancy is removed, and performs processing such as marking, correction or deletion according to the set rules. Set thresholds based on expert experience ,when , the corresponding data is marked as abnormal.

[0036] , , in, It indicates the degree of deviation of a certain data from the standard deviation of the data distribution after removing redundancy, and is used to detect abnormal data; is the mean of the data after removing redundancy; is the standard deviation of the data after removing redundancy; It is the data after anomaly detection.

[0037] Finally, the data after anomaly detection is stored in a predetermined format, an index is created in the database to optimize query efficiency, and the data after anomaly detection is stored in a suitable database.

[0038] , , in, For functions using AES-256 encryption, is the encrypted structured data. Indicates storing encrypted structured data in the database.

[0039] S2. Build a prediction model based on the encrypted structured data, generate prediction output, and use an adaptive incremental optimization algorithm to dynamically optimize and update the model parameters of the prediction model; use an improved relative error evaluation index to evaluate the prediction performance of the prediction model, and store the updated model parameters and prediction results in a database; In order to use the encrypted structured data To build a predictive model, we first need to decrypt the data and extract features. is the decrypted data, which is converted into a time series data matrix .

[0040] , in, is the time step; is the amount of decrypted data; is the index within the time window; Refers to the decrypted data In time Previously, by time step Backward offset The corresponding data at each time point.

[0041] Next, the phase space data is reconstructed by time delay embedding method .

[0042] , in, It's in time The decrypted data; It's in time The decrypted data; It's in time The decrypted data; It is a time delay; is the embedding dimension.

[0043] Then use the nonlinear phase space mapping function Modeling is done to adapt to changes in time series data.

[0044] , in, is the predicted output, indicating that The model predicted value of is the weight matrix; are the parameters in the weight matrix for linear transformation; is a bias term used to adjust the predicted output; Represents the activation function, which is used to introduce nonlinear characteristics; Each time new data is obtained Add the data matrix and update the phase space . New data points It means in time The collected and decrypted data will be used together with the historical data accumulated by the dispatching system for a long time to update the phase space matrix , for further predictive modeling.

[0045] The model parameters are dynamically optimized and updated using an adaptive incremental optimization algorithm by minimizing the increment of the prediction error.

[0046] , in, are the updated model parameters; It is An actual value, that is, the real target variable value; is the square norm of the weight matrix and the parameters in the weight matrix, used for regularization, For predictions using phase space data; Indicates time point; is the regularization coefficient, which is used to control the complexity of the model and avoid overfitting; is the number of actual values; is the actual value index.

[0047] Finally, the improved relative error evaluation index is used Evaluate the model prediction performance, store the updated model parameters and prediction results in the database, and generate real-time prediction charts for users to monitor and adjust.

[0048] , in, It is the predicted target variable value; is a small constant used to avoid division by zero errors.

[0049] S3. Based on the data and prediction output after anomaly detection, the task scheduling scheme is optimized through a dynamic adaptive fusion optimization algorithm; The data after anomaly detection obtained from S1 and the predicted output obtained from S2 Perform data fusion processing; , in, is the fused data; is the fusion weight, ranging from 0 to 1.

[0050] The scheduling objective function is used to comprehensively consider cost, completion time and resource utilization, and is defined by weighted summation.

[0051] , in, is the scheduling objective function; is the resource allocation variable; is the task scheduling variable; is the fused data; is the cost function; is the completion time function; is a resource utilization function; , , It is the weight coefficient of different optimization objectives in the scheduling objective function.

[0052] The cost function is calculated as: , in, It is The linear cost coefficient of each resource; It is The amount of resources allocated; Indicates Parameters related to the sensitivity of the allocation amount of each resource; Indicates Parameters related to the sensitivity of the allocation amount of each resource; Indicates Nonlinear cost coefficients related to the amount of resource allocation; It is with The offset associated with each resource allocation; is the total amount of resources; is the number of terms in the nonlinear cost calculation.

[0053] The calculation formula of the completion time function is: , in, It is The actual completion time of each task; is the average task completion time; It is a constant that prevents infinity or negative numbers from appearing in calculations; is the total number of tasks; Indicates the task index.

[0054] The resource utilization function calculation formula is: , in, is the fused data; is a positive definite matrix, representing the weight or correlation of resource allocation; is a constant that prevents infinite or negative numbers from appearing in calculations. yes The transpose of .

[0055] In order to dynamically adapt the fusion optimization algorithm, a dynamic adjustment mechanism is introduced; , in, is the dynamically adjusted weight coefficient; is the initial value of the weight coefficient; It is a parameter used to dynamically adjust the weights and is used to control the impact of real-time feedback errors on weight adjustment. is the real-time feedback error, which is used to reflect the deviation between the current system state and the expected target. It is a constant that prevents infinity or negative numbers from appearing in calculations.

[0056] Then use the dynamic adaptive fusion optimization algorithm to optimize the task scheduling scheme, the goal is to minimize the scheduling objective function The following formula is used to update resource allocation variables during optimization: and task scheduling variables : , , , in, is the updated resource allocation variable; is the updated task scheduling variable; is the resource allocation variable of the previous round; is the task scheduling variable of the previous round; is the learning rate, which is used to control the step size of parameter updates in the dynamic adaptive fusion optimization algorithm; , is the Jacobian matrix, which represents the sensitivity of the objective function to resource allocation and task scheduling; , is the objective function for resource allocation variables and task scheduling variables The gradient of , indicating the optimization direction; It is a constant that prevents infinity or negative numbers from appearing in calculations.

[0057] Through the above optimization process, the optimized scheduling objective function is obtained ; represents the optimized resource allocation variable, that is, the resource allocation variable after the final update; Represents the optimized task scheduling variable, that is, the task scheduling variable after the final update.

[0058] Furthermore, based on the expected value of the scheduling objective function predetermined by the scheduling system and the optimized scheduling objective function, the error is calculated to evaluate the performance of the scheduling system; the error calculation formula is as follows: , in, It's in time The error on is the optimized scheduling objective function; Represents the optimized resource allocation variable; Represents the optimized task scheduling variable; is the fused data; It is the expected value of the scheduling objective function predetermined by the scheduling system; It's in time Feedback error; It is a constant that prevents infinity or negative numbers from appearing in calculations; is the total number of tasks; Indicates the task index.

[0059] S4. Calculate performance evaluation indicators to evaluate the performance of the scheduling system in resource scheduling and task execution; generate optimization suggestions based on performance evaluation indicators; propose new resource allocation strategies and task scheduling solutions, and provide users with an interface for real-time monitoring of system performance and adjustment of system parameters, feedback prompts and suggestions, and historical data query functions.

[0060] Introduce performance evaluation indicators to evaluate the performance of the scheduling system in resource scheduling and task execution through the calculated errors. Generate a system performance evaluation report based on error analysis, including indicators such as average error, maximum error, and minimum error.

[0061] , , , in, is the average error, which indicates the average level of the overall performance of the scheduling system; is the maximum error, which indicates the maximum deviation of the scheduling system during the evaluation period and is used to identify extreme cases; is the minimum error, which indicates the minimum deviation of the scheduling system during the evaluation period and is used to identify the optimal performance point of the scheduling system; is the total time steps in the evaluation period; It's in time The error on .

[0062] Based on the performance evaluation indicators, optimization suggestions are automatically generated to improve the performance of the scheduling system. By analyzing the trend of error changes over time, potential optimization space can be identified. For example, the time period with high error occurrence and its corresponding system status can be identified.

[0063] Propose a new resource allocation strategy and task scheduling solution to optimize the scheduling system performance by reducing the load in the high error range. Adjust the learning rate based on error analysis , weight matrix The regularization parameter etc., to reduce overfitting or improve the generalization ability of the dynamic adaptive fusion optimization algorithm. Generate a report based on the error analysis results and optimization strategy for reference by the scheduling system administrator or user. The optimization strategy is a strategy based on error analysis, including new resource allocation strategy, adjustment of learning rate , adjust the regularization parameter wait.

[0064] The feedback data and performance evaluation indicators are then persistently stored, and the feedback data is encrypted; , in, It's in time The encrypted feedback data on is the encryption function; is the encryption key; It's in time Feedback data on Is the timestamp after synchronization.

[0065] The encrypted feedback data is stored in the database, and necessary indexes are created to optimize query efficiency for long-term monitoring and analysis. The stored historical feedback data is cleaned up regularly, and necessary feedback data is retained to support long-term trend analysis.

[0066] Provide users with interfaces including but not limited to the following for real-time monitoring of system performance and adjustment of system parameters.

[0067] Real-time monitoring interface: displays the current resource allocation, task scheduling, prediction results, error analysis and other information of the scheduling system; provides charts and dashboards to enable users to intuitively understand the real-time status of the scheduling system.

[0068] Parameter adjustment interface: provides users with the function of manually adjusting scheduling system parameters, such as weight coefficients, resource allocation strategies, etc., and realizes the switching between automatic and manual adjustment, so that users can temporarily take over the scheduling system under special circumstances.

[0069] Feedback tips and suggestions: Based on real-time monitored system performance data and all relevant data that affect scheduling and execution, such as sensor data, prediction data from prediction models, etc., optimization suggestions are automatically generated and displayed to remind users of potential performance issues and possible solutions.

[0070] Historical data query: Provides query functions for historical feedback data and performance evaluation reports, supporting users to review and analyze the long-term performance of the scheduling system.

[0071] In summary, a system and method for autonomous scheduling of disorderly planned grain storage and transportation is completed.

[0072] The order of the embodiments of the invention is for description only and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0073] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0074] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention.

Claims

1. A method for autonomously scheduling disordered grain storage and transportation plans, characterized in that: The following steps are involved: S1. Collect data in real time, calibrate and synchronize the data, and clean, filter, remove redundancy and detect anomalies on the synchronized data; Encrypt and structure the data after anomaly detection to obtain encrypted structured data, and store the encrypted structured data in the database; S2. Build a prediction model based on the encrypted structured data, generate prediction output, and use an adaptive incremental optimization algorithm to dynamically optimize and update the model parameters of the prediction model; use an improved relative error evaluation index to evaluate the prediction performance of the prediction model, and store the updated model parameters and prediction results in the database; S3. Based on the data and prediction output after anomaly detection, the task scheduling scheme is optimized through a dynamic adaptive fusion optimization algorithm; S4. Calculate performance evaluation indicators to evaluate the performance of the scheduling system in resource scheduling and task execution; generate optimization suggestions based on performance evaluation indicators; propose new resource allocation strategies and task scheduling solutions, and provide users with an interface for real-time monitoring of system performance and adjustment of system parameters, feedback prompts and suggestions, and historical data query functions.

2. The method for autonomous dispatching of disordered grain storage and transportation according to claim 1 is characterized in that: The S1 specifically includes: The collected data is transmitted to the central server or data receiving terminal, and the sensor output is initially calibrated using linear regression; the time synchronization algorithm is applied to adjust the timestamp of the data to match the server time, and the synchronized data is cleaned using the moving average algorithm to obtain smoothed data; the smoothed data is then filtered according to the set threshold range; redundancy is removed based on the filtered data, and duplicate data is identified and deleted; abnormal data in the data after redundancy is removed is identified through Z-score anomaly detection.

3. The method for autonomous dispatching of disorderly grain storage and transportation according to claim 1 is characterized in that: The S2 specifically includes: Decrypt the encrypted structured data to obtain the decrypted data, and convert the decrypted data into a time series data matrix; based on the decrypted data, reconstruct the phase space data through the time delay embedding method; then use the nonlinear phase space mapping function Construct a prediction model. The specific formula is: ， in, is the predicted output, indicating that The model predicted value of is the timestamp after synchronization; Represents phase space data; represents the nonlinear phase space mapping function; is the weight matrix; are the parameters in the weight matrix; is a bias term used to adjust the predicted output; Represents the activation function.

4. The method for autonomous dispatching of disorderly grain storage and transportation according to claim 3 is characterized in that: The S2 specifically includes: The specific formula of the adaptive incremental optimization algorithm is: ， in, are the model parameters of the updated prediction model; It is Actual values, representing the true target variable value; is the square norm of the weight matrix and the parameters in the weight matrix, used for regularization; For predictions using phase space data; Indicates time point; is the regularization coefficient; is the number of actual values; is the actual value index.

5. The method for autonomous dispatching of disorderly grain storage and transportation according to claim 1 is characterized in that: The S3 specifically includes: In the implementation process of the dynamic adaptive fusion optimization algorithm, data fusion processing is performed based on the abnormal detection data obtained from S1 and the prediction output obtained from S2 to obtain the fused data; and the scheduling objective function is defined by weighted summation based on the cost, completion time, resource utilization and fused data. The formula is as follows: ， in, is the scheduling objective function; is the resource allocation variable; is the task scheduling variable; is the fused data; is the cost function; is the completion time function; is a resource utilization function; , , It is the weight coefficient of different optimization objectives in the scheduling objective function.

6. The method for autonomous dispatching of disorderly grain storage and transportation according to claim 5 is characterized in that: The S3 specifically includes: In the process of implementing the dynamic adaptive fusion optimization algorithm, a dynamic adjustment mechanism is introduced to adjust the weight coefficient. The specific formula is: ， in, is the dynamically adjusted weight coefficient; is the initial value of the weight coefficient; It is a parameter used to dynamically adjust the weights and is used to control the impact of real-time feedback errors on weight adjustment. is the real-time feedback error; Indicates Parameters related to the sensitivity of the allocation amount of each resource; It is a constant that prevents infinity or negative numbers from appearing in calculations.

7. The method for autonomous dispatching of grain storage and transportation disorderly plan according to claim 5 is characterized in that: The S3 specifically includes: In the implementation process of the dynamic adaptive fusion optimization algorithm, the task scheduling scheme is optimized by minimizing the scheduling objective function and generating the optimized scheduling objective function; the scheduling system performance is evaluated by calculating the expected value of the scheduling objective function predetermined by the scheduling system and the error of the optimized scheduling objective function.

8. The method for autonomous dispatching of disorderly grain storage and transportation according to claim 7 is characterized in that: The S4 specifically includes: By analyzing the errors, the performance evaluation indicators are calculated and a system performance evaluation report is generated, including the average error, maximum error and minimum error.

9. The method for autonomous dispatching of disorderly grain storage and transportation according to claim 8 is characterized in that: The S4 specifically includes: Generate a report based on the error analysis results and optimization strategy for reference by the scheduling system administrator or user; the optimization strategy is a strategy formulated based on the error analysis, including a new resource allocation strategy, adjusting the learning rate, and adjusting the regularization parameter; the feedback data and performance evaluation indicators are then persistently stored, and the feedback data is encrypted; the encrypted feedback data is stored in the database and cleaned up regularly.

10. An autonomous dispatching system for grain storage and transportation disorderly plan, applied to the autonomous dispatching method for grain storage and transportation disorderly plan according to claim 1, characterized in that: Includes the following parts: Sensor network module, data acquisition and processing module, historical data analysis module, intelligent scheduling engine module, execution control module, feedback and optimization module, human-computer interaction interface module; The sensor network module is used to collect data on the environment, grain quality, equipment status and location in the warehouse and during transportation, perform automatic calibration and self-diagnosis in real time, add timestamps to all collected data for synchronization, and provide the synchronized data to the data collection and processing module; The data acquisition and processing module is used to receive the synchronized data from the sensor network module, and perform cleaning, screening, redundancy removal and anomaly detection processing. The data after anomaly detection is structured and stored in the database and provided to the historical data analysis module and the intelligent scheduling engine module; The historical data analysis module builds a prediction model based on the structured data after anomaly detection stored in the data acquisition and processing module, and dynamically updates and optimizes the prediction model using the historical data accumulated by the system over a long period of time, and provides the prediction results to the intelligent scheduling engine module; The intelligent scheduling engine module is used to receive the data after abnormal detection of the data acquisition and processing module, the prediction results of the historical data analysis module, the feedback information of the execution control module, the optimization suggestions and feedback data of the feedback and optimization module, and the adjustment instructions input by the user of the human-computer interaction interface module, and optimize the task scheduling plan by using the dynamic adaptive fusion optimization algorithm; and provide the optimized scheduling plan to the execution control module and the human-computer interaction interface module; The execution control module is used to receive the optimized scheduling plan of the intelligent scheduling engine module and the adjustment instructions input by the user of the human-computer interaction interface module, and convert them into execution commands and send them to the execution unit, monitor the execution of the commands in real time and identify abnormal situations, automatically adjust the execution commands, and provide the recorded command execution and abnormal report feedback information to the intelligent scheduling engine module, the feedback and optimization module and the human-computer interaction interface module; The feedback and optimization module is used to receive feedback information from the execution control module, compare and evaluate the expected and actual execution results, analyze the feedback information to generate optimization suggestions and store the feedback data persistently, and provide the optimization suggestions and feedback data to the intelligent scheduling engine module; The human-computer interaction interface module is used to receive the optimized scheduling plan of the intelligent scheduling engine module, the feedback information of the execution control module and the adjustment instructions input by the user, and provide a visual interface to display the system operation status, scheduling plan and execution status, allowing the user to monitor the system in real time, adjust parameters or perform emergency intervention, and provide the adjustment instructions input by the user to the intelligent scheduling engine module and the execution control module.

Images