Multi-machine collaborative automated operation method and system for train loading

Abstract

This invention relates to the field of multi-machine collaborative control technology, and particularly to a method and system for automated multi-machine collaborative operation of train loading. It includes: real-time acquisition of train location information and estimated arrival time fluctuation data through a sensor network, and obtaining a quantitative index of arrival time uncertainty using a time series prediction model; determining and prioritizing equipment allocation based on this index; adjusting equipment operating parameters and generating a preliminary adjustment plan by combining equipment status data and optimization algorithms; constructing a resource utilization model based on optimized parameters and multi-machine collaborative requirements to obtain a draft multi-machine collaborative rotation schedule, which is then refined to generate a refined schedule; and updating the equipment collaboration time based on this schedule and real-time feedback to obtain the final multi-machine collaborative scheduling scheme. This invention can dynamically adjust the operating status of loading equipment, improve the efficiency of equipment collaborative operation, reduce loading delays, and maximize resource utilization.

Description

Technical Field

[0001] This invention relates to the field of multi-machine collaborative control technology, and in particular to a multi-machine collaborative automated operation method and system for train loading. Background Technology

[0002] Traditional railway loading operations typically rely on manual scheduling and equipment operation. This not only increases labor costs but also struggles to cope with complex operating environments and fluctuations in train arrival times, leading to low loading efficiency, significant delays, and potentially resource waste and equipment conflicts. Therefore, there is an urgent need for a highly automated scheduling system capable of adapting to dynamic changes in real time to optimize the loading process and improve efficiency.

[0003] While existing technologies have proposed automated control-based loading and scheduling methods, most of these methods are limited to applications in static environments and lack the ability to dynamically adjust to fluctuations and uncertainties in train arrival times and the collaborative operation of equipment. Furthermore, most existing technologies rely on simple scheduling algorithms and fail to fully utilize real-time data for optimization, resulting in low scheduling efficiency in complex scenarios.

[0004] This invention provides an automated operation method and system based on multi-machine collaboration. It collects real-time train location information and estimated arrival time fluctuation data on railway tracks, and analyzes this data using an advanced time series prediction model to obtain a quantitative index of the uncertainty of train arrival time. Based on this index, the allocation priority and operating parameters of loading equipment are dynamically adjusted. According to real-time feedback data, the collaboration time and operating status of the equipment are adjusted to ensure efficient collaborative operation between the equipment.

[0005] This invention utilizes a multi-machine collaborative scheduling method based on real-time optimization feedback and resource utilization models. It can adjust the operating strategy of loading equipment in a timely manner according to the dynamic changes in train arrival times, maximizing equipment resource utilization, minimizing loading delays, and improving the automation level and operational efficiency of railway loading operations. This not only optimizes the railway transportation workflow but also provides a new solution for intelligent scheduling of railway loading. It can adapt to complex railway environments and uncertain operating conditions, solving problems such as low efficiency, slow response, and resource waste caused by traditional manual scheduling methods. Summary of the Invention

[0006] This invention provides a multi-machine collaborative automated operation method and system for train loading, which maximizes resource utilization, reduces loading delays, and ensures efficient operation in complex and dynamic railway environments by dynamically adjusting the operating parameters and coordinating the scheduling of loading equipment. It can effectively cope with the challenges of uncertain train arrival times and equipment scheduling.

[0007] In a first aspect, the present invention provides a multi-machine collaborative automated operation method for loading trains, the method comprising:

[0008] Step S1: Obtain real-time location data and estimated arrival time fluctuation information of trains on railway tracks through sensor networks, process the real-time location data and estimated arrival time fluctuation information using a time series prediction model, and obtain a quantitative index of arrival time uncertainty.

[0009] Step S2: Based on the arrival time uncertainty quantification index, determine the relationship between the quantification index and the preset threshold, and determine the allocation priority sequence of loading equipment;

[0010] Step S3: Obtain the allocation priority sequence and the current loading equipment status data, use an optimization algorithm to adjust the equipment operating parameters to obtain a preliminary adjustment plan; based on the preliminary adjustment plan, verify whether the preliminary adjustment plan meets the preset requirements to obtain the optimized parameters;

[0011] Step S4: Based on the optimization parameters and multi-machine collaboration requirements, construct a resource utilization model to obtain a draft multi-machine collaboration rotation schedule; based on the draft rotation schedule, determine the relationship between the decision time of the draft and the preset time limit to obtain the final refined schedule;

[0012] Step S5: Based on the final refining schedule and real-time optimization feedback, update the collaboration time between devices to obtain a multi-machine collaborative scheduling scheme.

[0013] As a preferred embodiment of the present invention, step S1 includes:

[0014] Real-time location data and estimated arrival time fluctuation information of trains on railway tracks are collected through a sensor network; the real-time location data is preprocessed to obtain a first location dataset; features are extracted from the estimated arrival time fluctuation information to obtain a time fluctuation feature set; the first location dataset and the time fluctuation feature set are input into a pre-trained time series prediction model to obtain a first prediction result; based on the first prediction result, an arrival time uncertainty quantification index is calculated, wherein the quantification index characterizes the fluctuation range of train arrival time; and the arrival time uncertainty quantification index is output.

[0015] As a preferred embodiment of the present invention, step S2 includes:

[0016] Determine whether the quantification index is greater than a preset threshold. If so, mark it as a high-dynamic scenario; otherwise, mark it as a low-dynamic scenario. Determine a first priority sequence based on the high-dynamic scenario or the low-dynamic scenario, wherein the first priority sequence includes the order of device allocation. Sort the first priority sequence to obtain a second priority sequence. Output the second priority sequence as the device allocation priority sequence.

[0017] As a preferred embodiment of the present invention, step S3 yields a preliminary adjustment scheme, including:

[0018] Obtain the allocation priority sequence and the current loading equipment status data; input the allocation priority sequence and the current loading equipment status data into a pre-established genetic algorithm model; calculate the pitch angle parameters of the bucket wheel reclaimer and the speed parameters of the belt conveyor through the genetic algorithm model to obtain the first parameter combination; the genetic algorithm model adopts a real number encoded genetic algorithm.

[0019] The fitness of the first parameter combination is evaluated to obtain a first fitness value; based on the first fitness value, the parameters in the genetic algorithm model are iteratively adjusted to obtain a second parameter combination, which serves as a preliminary adjustment scheme.

[0020] As a preferred embodiment of the present invention, in step S3, based on the preliminary adjustment scheme, whether the preliminary adjustment scheme meets the preset requirements is verified to obtain the optimized parameters, including:

[0021] Extract a first parameter combination from the preliminary adjustment scheme; obtain a preset emergency demand response attribute; determine whether the first parameter combination satisfies the emergency demand response attribute. If it does, determine the first parameter combination as the optimization parameter; if it does not, iteratively adjust the genetic algorithm model, incorporate dynamic fusion factors, and obtain a second parameter combination; verify the second parameter combination to obtain a third parameter combination, which is then used as the optimization parameter.

[0022] As a preferred embodiment of the present invention, step S4 involves obtaining a draft multi-machine collaborative rotation schedule, including:

[0023] Obtain the optimization parameters and multi-machine collaboration requirements; determine the operating status parameters of each device based on the optimization parameters; determine the collaboration constraints between devices based on the multi-machine collaboration requirements; input the operating status parameters and collaboration constraints into a pre-established resource utilization model; calculate the rotation interval between devices using the resource utilization model to obtain a first rotation schedule; perform constraint verification on the first rotation schedule to obtain a second rotation schedule; output the second rotation schedule as a draft multi-machine collaborative rotation schedule.

[0024] As a preferred embodiment of the present invention, step S4, obtaining the final refining time schedule, includes:

[0025] Obtain the draft rotation schedule; calculate the decision time of the draft rotation schedule; if the decision time is greater than a preset time limit, use a parallel computing method to accelerate the parameter calculation process and obtain a first refined schedule; if the decision time is less than or equal to the preset time limit, determine the draft rotation schedule as the first refined schedule; perform constraint optimization on the first refined schedule to obtain a second refined schedule; output the second refined schedule as the final refined schedule.

[0026] As a preferred embodiment of the present invention, step S5, obtaining a multi-machine collaborative scheduling scheme, includes:

[0027] The process involves: acquiring the final refining schedule and real-time optimization feedback data; determining a first adjustment parameter for the equipment allocation status based on the real-time optimization feedback data; calculating a second adjustment parameter for the equipment allocation status based on the final refining schedule; inputting the first and second adjustment parameters into a pre-established status update model; calculating the updated value of the equipment allocation status through the status update model to obtain a first scheduling scheme; performing target optimization on the first scheduling scheme to obtain a second scheduling scheme, wherein the target optimization includes minimizing loading delays; and using the second scheduling scheme as a multi-machine collaborative scheduling scheme.

[0028] Secondly, the present invention also provides a multi-machine collaborative automated operation system for train loading, for implementing the above-mentioned method, the system comprising:

[0029] The data acquisition unit is used to acquire real-time location data and estimated arrival time fluctuation information of trains on railway tracks through sensor networks, and to process the real-time location data and estimated arrival time fluctuation information using a time series prediction model to obtain a quantitative index of arrival time uncertainty.

[0030] The priority judgment unit is used to determine the relationship between the arrival time uncertainty quantification index and the preset threshold, and to determine the allocation priority sequence of the loading equipment;

[0031] The scheme optimization unit is used to acquire the allocation priority sequence and the current loading equipment status data, adjust the equipment operating parameters using an optimization algorithm to obtain a preliminary adjustment scheme; and verify whether the preliminary adjustment scheme meets the preset requirements to obtain the optimization parameters.

[0032] The model building unit is used to construct a resource utilization model based on the optimization parameters and multi-machine collaboration requirements to obtain a draft multi-machine collaboration rotation schedule; based on the draft rotation schedule, it determines the relationship between the decision time of the draft and the preset time limit to obtain the final refined schedule;

[0033] The scheduling scheme update unit is used to update the collaboration time between devices based on the final refining schedule and real-time optimization feedback, so as to obtain a multi-machine collaborative scheduling scheme.

[0034] Thirdly, the present invention also provides a computer-readable storage medium storing instructions that, when executed by a processor, implement the above-described method.

[0035] The beneficial effects of this invention are as follows:

[0036] This invention acquires real-time train location information and estimated arrival time fluctuation data through a sensor network. Using a time series prediction model, it obtains a quantitative index of train arrival time uncertainty. Based on this index, it compares it with a preset threshold to determine if the current scenario is high-dynamic, and generates a priority sequence for loading equipment allocation. In high-dynamic scenarios, equipment is prioritized for allocation to track sections with larger fluctuations to cope with sudden changes in arrival time. Based on the priority sequence and equipment status data, an optimization algorithm adjusts the equipment's operating parameters to obtain a preliminary equipment parameter adjustment plan. After verification, this plan ensures that the preset requirements for loading operations are met, thereby generating optimized parameters. This allows the equipment to adjust its operating status promptly and accurately in dynamic scenarios to adapt to constantly changing operational needs. Furthermore, it utilizes multi-machine collaboration... The requirements include constructing a resource utilization model and generating a draft multi-machine collaborative rotation schedule to ensure that the rotation interval matches the loading operation time requirements. If the decision-making time exceeds the preset limit, parallel computing methods are employed to further improve decision-making efficiency. Ultimately, the refined schedule provides a clear time arrangement for subsequent equipment collaborative operations. Based on the final refined schedule and real-time optimization feedback, the collaboration time between equipment is updated to generate a multi-machine collaborative scheduling scheme. This scheme can adjust the collaborative status of equipment in real time, ensuring efficient equipment collaboration and reducing resource waste and delays caused by equipment idleness or mismatched work. Through the mutual cooperation of the above technical solutions, it is ensured that equipment can flexibly adjust and collaborate in dynamic railway loading scenarios, thereby effectively improving operational efficiency, reducing delays, and providing technical support for the automation and intelligence of railway loading operations. Attached Figure Description

[0037] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 This is a flowchart of a multi-machine collaborative automated operation method for loading trains, as shown in the embodiment.

[0039] Figure 2 This is a flowchart of the method for obtaining the preliminary adjustment scheme in the embodiment;

[0040] Figure 3 This is a flowchart illustrating the method for obtaining optimization parameters in this embodiment;

[0041] Figure 4 This is a structural diagram of a multi-machine collaborative automated operation system for loading trains, as shown in the embodiment. Detailed Implementation

[0042] This invention provides a method and system for multi-machine collaborative automated operation of train loading. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or device that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0043] For ease of understanding, the specific process of the embodiments of the present invention will be described below, such as... Figure 1 As shown, the multi-machine collaborative automated operation method for train loading in this embodiment of the invention specifically includes:

[0044] Step S1: Acquire real-time location data and estimated arrival time fluctuation information of trains on the railway track through a sensor network, and process the real-time location data and estimated arrival time fluctuation information using a time series prediction model to obtain a quantitative index of arrival time uncertainty; specifically including:

[0045] Real-time location data and estimated arrival time fluctuation information of trains on railway tracks are collected through a sensor network; the real-time location data is preprocessed to obtain a first location dataset; features are extracted from the estimated arrival time fluctuation information to obtain a time fluctuation feature set; the first location dataset and the time fluctuation feature set are input into a pre-trained time series prediction model to obtain a first prediction result; based on the first prediction result, an arrival time uncertainty quantification index is calculated, wherein the quantification index characterizes the fluctuation range of train arrival time; and the arrival time uncertainty quantification index is output.

[0046] Specifically, a sensor network is used to collect real-time location data and estimated arrival time fluctuations of trains on railway tracks. Sensor nodes deployed along the railway line monitor the train's GPS coordinates and speed signals in real time, capturing estimated arrival time deviations. The sensor nodes transmit the collected information via a wireless sensor network, ensuring accurate and efficient data transmission to the processing system. After acquiring the real-time location data, preprocessing is performed, primarily including noise filtering and missing value imputation, to ensure data integrity and reliability. This preprocessed data serves as the first location dataset for subsequent analysis and prediction. For the estimated arrival time fluctuations, feature extraction is performed by statistically analyzing the differences between historical and actual arrival times, resulting in a time fluctuation feature set. Statistical processing of the fluctuation information extracts features such as mean, variance, and peak value. These features reflect changes in arrival time and provide input for subsequent time series prediction. During feature extraction, the extracted features are standardized to adapt to the input requirements of the prediction model. The purpose of feature extraction is to capture the dynamic changes in train arrival times, especially during peak railway scheduling periods, effectively identifying potential delays. The pattern is used to improve the accuracy of subsequent predictions. The preprocessed first position dataset and the extracted time fluctuation feature set are input into a pre-trained time series prediction model to obtain the first prediction result. The time series prediction model used, such as a Long Short-Term Memory (LSTM) network, is trained on historical data, which can capture the temporal dependence of the data and effectively handle the gradient vanishing problem in the data. The historical data includes the historical first position dataset, the corresponding time fluctuation features, and the actual arrival time. The first position dataset and the time fluctuation feature set are combined into an input sequence, that is, the above first... In the location dataset, the train's position, speed, and arrival time fluctuation features at each moment are combined into an input vector, which is used as the model input for forward propagation to obtain the predicted arrival time series. Since the historical arrival time of a train is not exactly the same each time, the model's output is a prediction of the arrival time for future time periods, i.e., the predicted arrival time range. Therefore, the above first prediction result provides an estimate of the train's arrival time and reveals the possible range of arrival time fluctuations. For example, the first location dataset is the preprocessed real-time location data. After preprocessing, the data at each timestamp contains the train's precise location and speed information. For example:

[0047] Timestamp T0 (e.g., 10:00:00): Latitude and Longitude (118.12345, 24.56789), Speed ​​80km / h

[0048] Timestamp T1 (10:01:00): Latitude and longitude (118.12500, 24.56900), speed 82km / h

[0049] Timestamp T2 (10:02:00): Latitude and longitude (118.12650, 24.57010), speed 79km / h

[0050] The time fluctuation feature set refers to the statistical features extracted from the expected arrival time fluctuation information, and from historical data and real-time information, that characterize the stability of arrival time. For example, for the same point in time,

[0051] At time T0: the mean deviation of the predicted arrival time over the past 5 minutes is -0.5 minutes, the variance is 1.2 minutes, and the peak value of the predicted change in the most recent minute is 0.8 minutes; at time T1: the mean deviation is +0.2 minutes, the variance is 1.5 minutes, and the peak value is 0.6 minutes; at time T2: the mean deviation is -0.1 minutes, the variance is 1.0 minute, and the peak value is 0.3 minutes; for each time point, the location data (longitude, latitude, speed) and the time fluctuation characteristics (mean deviation, variance, peak value) are combined into a comprehensive vector;

[0052] Input vector at time T0: [118.12345, 24.56789, 80, -0.5, 1.2, 0.8]

[0053] The input vector at time T1 is: [118.12500, 24.56900, 82, +0.2, 1.5, 0.6].

[0054] The input vector at time T2 is: [118.12650,24.57010,79,-0.1,1.0,0.3].

[0055] Based on the first prediction result, an uncertainty quantification index for arrival time is calculated. This uncertainty index characterizes the fluctuation range of train arrival time. Specifically, confidence intervals of the prediction sequence are extracted from the first prediction result, and multiple prediction samples are generated using the Monte Carlo dropout method. These samples represent the uncertainty of the prediction result. By calculating the statistical variance of these samples, the uncertainty quantification index can be obtained. This index reflects the possible fluctuation range of arrival time in the form of standard deviation. By applying confidence levels to the prediction result, such as a 95% confidence interval, the upper and lower limits of the fluctuation range can be quantified, thereby clarifying the reliability and fluctuation amplitude of the prediction result. To ensure that the uncertainty quantification indicators are closer to real-world scenarios, historical fluctuation data is incorporated to adjust the indicators and better reflect the actual transportation environment. Ultimately, the calculated arrival time uncertainty quantification indicators are transmitted to the downstream scheduling module via an interface for subsequent decision-making. The output of these uncertainty quantification indicators ensures that the scheduling system can obtain real-time dynamic changes in train arrival times, enabling precise equipment allocation and task scheduling. This not only enhances the dynamic optimization capabilities of the railway loading process but also reduces resource waste and loading delays caused by arrival time fluctuations, significantly improving the efficiency and accuracy of multi-machine collaborative operations during railway loading.

[0056] Step S2: Based on the arrival time uncertainty quantification index, determine the relationship between the quantification index and a preset threshold, and determine the allocation priority sequence of loading equipment; specifically including:

[0057] Determine whether the quantification index is greater than a preset threshold. If so, mark it as a high-dynamic scenario; otherwise, mark it as a low-dynamic scenario. Determine a first priority sequence based on the high-dynamic scenario or the low-dynamic scenario, wherein the first priority sequence includes the order of device allocation. Sort the first priority sequence to obtain a second priority sequence. Output the second priority sequence as the device allocation priority sequence.

[0058] Specifically, the system determines whether the aforementioned arrival time uncertainty quantification index exceeds a preset threshold. If the quantification index exceeds the preset threshold, it is marked as a high-dynamic scenario, indicating that the arrival time fluctuates significantly and equipment needs to be prioritized for scheduling to address the impact of this uncertainty. Conversely, if the quantification index is less than or equal to the preset threshold, it is marked as a low-dynamic scenario, indicating that the arrival time is relatively stable and equipment allocation can proceed in the normal order. By judging the magnitude of the quantification index, the system can dynamically divide scenarios and adjust the priority of equipment allocation according to the characteristics of different scenarios. Based on the aforementioned high-dynamic or low-dynamic scenarios, a first priority sequence is determined, which includes the order of equipment allocation. Specifically, in high-dynamic scenarios, equipment such as bucket wheel reclaimers and belt conveyors are prioritized for allocation to track sections with greater fluctuations. The equipment allocation order is sorted from high to low based on the magnitude of the uncertainty index to ensure timely response to high-fluctuation areas. In low-dynamic scenarios, the equipment allocation order is arranged based on equipment availability or other static factors, such as equipment status and historical performance, to ensure the overall stability of scheduling.

[0059] After determining the first priority sequence, the sequence is further sorted using optimization algorithms such as genetic algorithms to generate a second priority sequence. The genetic algorithm is a real-number encoded genetic algorithm that simulates the natural selection process, optimizing the priority order of equipment through iterative selection, crossover, and mutation operations. It also considers current equipment status data, such as the pitch angle of the bucket wheel reclaimer and the speed of the belt conveyor, to ensure the optimized priority sequence meets the goal of minimizing loading delays. During the sorting process, the genetic algorithm also incorporates emergency demand response attributes. If some equipment cannot meet specific needs, parameters are dynamically adjusted to ensure that equipment allocation can effectively cope with fluctuations in arrival time in highly dynamic scenarios. Finally, the system outputs the optimized second priority sequence as the equipment allocation priority sequence. This allocation priority sequence is used to update the equipment allocation status in real time to meet the goal of minimizing loading delays and optimize resource utilization efficiency through a multi-machine collaborative scheduling scheme. This technical solution, through precise scenario division and priority sorting, can significantly improve the scheduling efficiency during train loading, reduce delays caused by uncertainty in arrival time, and ensure the efficiency and smoothness of railway transportation.

[0060] Step S3: Obtain the allocation priority sequence and the current loading equipment status data, use an optimization algorithm to adjust the equipment operating parameters to obtain a preliminary adjustment plan; based on the preliminary adjustment plan, verify whether the preliminary adjustment plan meets the preset requirements to obtain the optimized parameters;

[0061] In step S3, a preliminary adjustment plan is obtained, such as... Figure 2 As shown, it includes:

[0062] Obtain the allocation priority sequence and the current loading equipment status data; input the allocation priority sequence and the current loading equipment status data into a pre-established genetic algorithm model; calculate the pitch angle parameters of the bucket wheel reclaimer and the speed parameters of the belt conveyor through the genetic algorithm model to obtain the first parameter combination; the genetic algorithm model adopts a real number encoded genetic algorithm.

[0063] The fitness of the first parameter combination is evaluated to obtain a first fitness value; based on the first fitness value, the parameters in the genetic algorithm model are iteratively adjusted to obtain a second parameter combination, which serves as a preliminary adjustment scheme.

[0064] Specifically, the allocation priority sequence and current loading equipment status data are obtained. The allocation priority sequence is calculated based on the arrival time uncertainty quantification index. Changes in the quantification index determine whether the current scenario is high-dynamic or low-dynamic. The current loading equipment status data includes parameters such as the pitch angle of the bucket wheel reclaimer and the speed of the belt conveyor, reflecting the current working status of the loading equipment. The allocation priority sequence and current loading equipment status data are input into a pre-established genetic algorithm model. The genetic algorithm is an optimization algorithm that simulates natural selection and genetic mechanisms, searching for the optimal solution through selection, crossover, and mutation operations. In the genetic algorithm model, the allocation priority sequence serves as the initial population condition input, while the loading equipment status data is used to define the constraint boundaries in the genetic algorithm. The model calculates the pitch angle of the bucket wheel reclaimer... Angle parameters and conveyor speed parameters are used to generate preliminary parameter combinations. For example, in highly dynamic scenarios, the genetic algorithm considers that adjustments to loading equipment parameters will prioritize responses to track segments with large fluctuations in arrival time. After obtaining the preliminary parameter combinations, a fitness evaluation is performed on these combinations, calculating their fitness values. The goal of the fitness evaluation is to determine the merits of the combinations by quantifying the weighted sum of loading efficiency and delay minimization. Through simulation calculations, the efficiency and delay of each combination are evaluated, and a first fitness value is output. The evaluation results are used for subsequent iterative adjustments. During the iteration process, if the first fitness value is lower than a set threshold, the genetic algorithm model will iterate by increasing the mutation rate or adjusting the selection and crossover parameters to generate new parameter combinations. After each iteration, the new parameter combinations will be re-evaluated until the fitness value exceeds the preset threshold, thus yielding a second parameter combination.

[0065] The aforementioned second parameter combination serves as a preliminary adjustment scheme, which will be used to adjust equipment parameters to optimize equipment collaboration during the loading process. A genetic algorithm ensures precise adjustment of equipment parameters, enabling rapid adaptation to changes in highly dynamic or low-dynamic scenarios, thus improving overall scheduling efficiency. The generated preliminary adjustment scheme will undergo further verification and optimization to ultimately form a scheduling scheme that meets the needs of multi-machine collaboration, thereby achieving dynamic optimization during railway loading, improving resource utilization, and reducing loading delays. This technical solution, by acquiring real-time equipment status and scenario dynamic information, combined with genetic algorithms for parameter optimization and adjustment, ensures efficient operation and rational allocation of loading equipment in complex dynamic environments. This not only improves the flexibility and accuracy of scheduling but also effectively enhances the automation level of the train loading process.

[0066] Further, in step S3, based on the preliminary adjustment plan, it is verified whether the preliminary adjustment plan meets the preset requirements, and optimized parameters are obtained, such as... Figure 3 As shown, it includes:

[0067] Extract a first parameter combination from the preliminary adjustment scheme; obtain a preset emergency demand response attribute; determine whether the first parameter combination satisfies the emergency demand response attribute. If it does, determine the first parameter combination as the optimization parameter; if it does not, iteratively adjust the genetic algorithm model, incorporate dynamic fusion factors, and obtain a second parameter combination; verify the second parameter combination to obtain a third parameter combination, which is then used as the optimization parameter.

[0068] Specifically, in this embodiment, the first parameter combination extracted from the preliminary adjustment scheme includes equipment parameters optimized by a genetic algorithm, such as the pitch angle of the bucket wheel reclaimer and the speed of the belt conveyor. The preliminary adjustment scheme is an initial set of parameters obtained by optimizing the equipment parameters using a genetic algorithm after considering the dynamic requirements of different scenarios. Preset emergency demand response attributes are also obtained, including a train arrival time fluctuation threshold and an equipment response time limit, used to evaluate whether the parameter combination can meet the requirements under highly dynamic scenarios. First, it is determined whether the extracted first parameter combination meets the aforementioned emergency demand response attributes. If the first parameter combination can meet these attributes, for example, if its expected response time is less than the preset response time limit... The first parameter combination is directly determined as the optimization parameter, which can quickly respond to fluctuations during loading, thereby reducing delays and improving efficiency. If the first parameter combination does not meet the required emergency response attributes, it enters the iterative adjustment stage, specifically readjusting the genetic algorithm to incorporate dynamic fusion factors to cope with uncertainties. For example, it identifies the shortcomings of the first parameter combination, such as the response delay caused by the pitch angle parameter exceeding a preset threshold. Then, by adding dynamic fusion factors, such as the volatility of real-time train position data, it adjusts the parameters such as population selection and crossover probability in the genetic algorithm, making the genetic algorithm biased towards equipment parameter combinations that can quickly respond to fluctuations, and gradually generates the second parameter combination, thereby effectively improving the adaptability of the genetic algorithm to highly dynamic scenarios and reducing loading delays.

[0069] The iteratively adjusted second parameter combination underwent further verification to confirm its compliance with all requirements. Specifically, this was achieved by simulating resource utilization under multi-machine collaboration, particularly by calculating the rotation interval of loading equipment, to confirm whether it met the emergency demand response attributes. If the simulation results showed that the rotation interval exceeded the preset limit, the parameters were fine-tuned, such as adjusting the belt conveyor speed, until the rotation interval met the requirements. This ensured the robustness of the second parameter combination and prevented equipment failure or untimely response due to inappropriate parameters in practical applications. The adjusted second parameter combination was then used as the third parameter combination, i.e., the optimized parameters, and further used to construct a resource utilization model, thereby generating a draft multi-machine collaborative rotation schedule. These optimized parameters ensured that equipment could effectively collaborate in complex dynamic environments and respond promptly to changing demands in high-dynamic or low-dynamic scenarios. By combining the flexibility of genetic algorithms with real-time data feedback, not only was the efficiency of multi-machine collaborative scheduling improved, but loading delays caused by fluctuations in train arrival times were also significantly reduced, ensuring high efficiency and stability in railway transportation.

[0070] Step S4: Based on the optimization parameters and multi-machine collaboration requirements, construct a resource utilization model to obtain a draft multi-machine collaboration rotation schedule; based on the draft rotation schedule, determine the relationship between the decision time of the draft and the preset time limit to obtain the final refined schedule;

[0071] In step S4, a draft multi-machine collaborative rotation schedule is obtained, including:

[0072] Obtain the optimization parameters and multi-machine collaboration requirements; determine the operating status parameters of each device based on the optimization parameters; determine the collaboration constraints between devices based on the multi-machine collaboration requirements; input the operating status parameters and collaboration constraints into a pre-established resource utilization model; calculate the rotation interval between devices using the resource utilization model to obtain a first rotation schedule; perform constraint verification on the first rotation schedule to obtain a second rotation schedule; output the second rotation schedule as a draft multi-machine collaborative rotation schedule.

[0073] Specifically, in this embodiment, optimization parameters and multi-machine coordination requirements are first obtained. The optimization parameters include combinations of equipment parameters obtained through a genetic algorithm, such as the pitch angle of the bucket wheel reclaimer and the speed of the belt conveyor. The multi-machine coordination requirements include a lower limit for equipment rotation intervals and the goal of minimizing loading delays. To ensure the accuracy of subsequent calculations, relevant data is collected in real time through a sensor network, ensuring that the optimization process can be adjusted according to actual conditions. After obtaining the optimization parameters, the operating status parameters of each device are determined based on these parameters. For example, the material handling efficiency parameter of the bucket wheel reclaimer is calculated based on its pitch angle, and the material flow is deduced by combining this with the speed of the belt conveyor. The operating rate is used to obtain the load status parameters of each device. These operating status parameters can accurately describe the working status of the loading equipment, providing a basis for subsequent resource allocation and coordinated scheduling. Based on the above multi-machine coordination requirements, the coordination constraints between loading equipment are further determined. For example, the coordination requirements include the synchronous material transfer requirements between the bucket wheel reclaimer and the belt conveyor. Therefore, constraints such as the rotation interval not being less than a specific time value are set to avoid equipment conflicts and ensure the stable operation of the system. In addition, dynamic constraints are set according to the uncertainty and expected fluctuations of train arrival time to ensure that it can adapt to different changing needs in highly dynamic scenarios.

[0074] After inputting the operating status parameters and collaborative constraints of the loading equipment into a pre-established resource utilization model, the rotation interval between each piece of equipment is calculated by solving the resource utilization model, thus obtaining the first rotation schedule. The resource utilization model employs a linear programming-based mathematical framework, optimizing the objective function, such as minimizing loading delays, to achieve efficient equipment scheduling. During model calculation, operating status parameters, such as load status, are used as variable inputs, and collaborative constraints are transformed into inequality restrictions, ensuring that all constraints are considered during the calculation. Furthermore, for highly dynamic scenarios, the resource utilization model also incorporates time series forecasting results, enabling the model to adjust parameters in real time to better adapt to the constantly changing railway environment. After calculating the first rotation schedule, constraint verification is performed to ensure it complies with... All collaborative constraints are considered. By comparing the rotation intervals in the first rotation schedule with the preset constraints item by item, if a violation is found, the interval value is incremented and adjusted until the schedule meets all constraints, thus generating the second rotation schedule. The above verification process ensures the feasibility and efficiency of the rotation schedule, avoiding scheduling conflicts or resource waste caused by non-compliance with constraints. The verified and adjusted second rotation schedule is output as a draft of the multi-machine collaborative rotation schedule. The second rotation schedule lists the specific rotation periods of each device and provides basic data for the subsequent refinement process. Through the above technical solution, an effective multi-machine collaborative scheduling scheme that can adapt to dynamic changes can be generated based on real-time data and multi-machine collaborative requirements, minimizing loading delays, optimizing resource utilization, and ensuring the efficiency and stability of the railway loading process.

[0075] Further, in step S4, the final refining schedule is obtained, including:

[0076] Obtain the draft rotation schedule; calculate the decision time of the draft rotation schedule; if the decision time is greater than a preset time limit, use a parallel computing method to accelerate the parameter calculation process and obtain a first refined schedule; if the decision time is less than or equal to the preset time limit, determine the draft rotation schedule as the first refined schedule; perform constraint optimization on the first refined schedule to obtain a second refined schedule; output the second refined schedule as the final refined schedule.

[0077] Specifically, in this embodiment, firstly, a draft rotation schedule is obtained. This draft schedule is generated based on optimized parameters and multi-machine collaboration requirements, specifically including the rotation intervals between each piece of equipment calculated using a resource utilization model. The draft rotation schedule includes preliminary rotation arrangements for the bucket wheel reclaimer and belt conveyor. The decision time for the draft rotation schedule is also calculated, specifically including the cumulative time spent calculating all parameters in the draft, for example, the total time from the parameter combination optimized by the genetic algorithm to the calculation of the rotation interval. This helps quantify the efficiency of the decision-making process and ensure the accuracy of subsequent judgments. If the calculated decision time exceeds a preset time limit, the parameter calculation process is accelerated using parallel computing methods to obtain a first refined schedule. The specific decision time... The timetable is compared with a preset time limit. If the limit is exceeded, parallel computing is initiated. The parameter calculation process is decomposed into multiple sub-tasks, such as parallel threads for pitch angle optimization and rate adjustment. These tasks are executed simultaneously using multi-core processors, which greatly shortens the computation time and ensures that the refined timetable can be updated in a timely manner in highly dynamic scenarios. For example, in scenarios with high uncertainty in train arrival time, parallel computing can significantly reduce decision latency, thereby accelerating the update of the device allocation priority sequence. If the decision time is less than or equal to the preset time limit, no accelerated computation is required. The draft rotation timetable is directly determined as the first refined timetable without additional acceleration processing. At this time, the draft rotation timetable can meet the real-time requirements, so the constraint optimization process can be directly entered.

[0078] The first refined timetable is further optimized by constraints to obtain the second refined timetable. Linear programming is applied to impose constraints on the rotation intervals in the timetable, such as considering limitations imposed by equipment status data and emergency demand response attributes. By iteratively adjusting parameter combinations, loading delays are minimized. By minimizing the total rotation time objective function, the collaboration between loading equipment is optimized, improving overall scheduling efficiency. For example, in low-dynamic scenarios, constraint optimization adjusts the rotation intervals from their initial values ​​to shorter times, thereby reducing equipment idle time. In high-dynamic scenarios, the rotation intervals are further optimized to cope with fluctuations in train position data, thus avoiding delay accumulation. Finally, the second refined timetable, after constraint optimization, is output as the final refined timetable. This final refined timetable is transmitted to the loading equipment allocation status update module, becoming part of the final multi-machine collaborative scheduling scheme, guiding subsequent operations and decisions. This technical solution, through a combination of real-time computation, parallel processing, and constraint optimization, ensures that an efficient and accurate multi-machine collaborative rotation timetable can be quickly obtained in complex dynamic scenarios, effectively improving resource utilization efficiency and scheduling response speed in the railway loading process.

[0079] Step S5: Based on the final refining schedule and real-time optimization feedback, update the collaboration time between devices to obtain a multi-machine collaborative scheduling scheme; specifically including:

[0080] The process involves: acquiring the final refining schedule and real-time optimization feedback data; determining a first adjustment parameter for the equipment allocation status based on the real-time optimization feedback data; calculating a second adjustment parameter for the equipment allocation status based on the final refining schedule; inputting the first and second adjustment parameters into a pre-established status update model; calculating the updated value of the equipment allocation status through the status update model to obtain a first scheduling scheme; performing target optimization on the first scheduling scheme to obtain a second scheduling scheme, wherein the target optimization includes minimizing loading delays; and using the second scheduling scheme as a multi-machine collaborative scheduling scheme.

[0081] Specifically, in this embodiment, a multi-machine collaborative scheduling scheme is first generated by updating the collaboration time between loading equipment based on the final refining schedule and real-time optimization feedback data. Specifically, the final refining schedule and real-time optimization feedback data are first acquired. The real-time optimization feedback data includes fluctuation information on train arrival times. This fluctuation information reflects the dynamics of the current scenario, thus providing a basis for subsequent equipment scheduling. Based on this data, a first adjustment parameter for the equipment allocation status is further determined. Specifically, the first adjustment parameter is an adjustment value calculated by applying a linear regression method to process the arrival time uncertainty quantification index in the real-time optimization feedback data. The linear regression model uses the uncertainty index as the independent variable and the equipment status deviation as the dependent variable, and obtains the adjustment coefficient through least squares fitting to form the first adjustment parameter. Based on the final refining schedule, a second adjustment parameter for the equipment allocation status is calculated. This involves extracting rotation interval data from the refining schedule and calculating the second adjustment parameter based on the rotation interval data, which serves as a correction value for the equipment status according to the schedule. The first and second adjustment parameters are then input into a pre-established status update model to obtain updated equipment status values. This status update model is a dynamic model based on a Kalman filter. The model is trained using a machine learning algorithm based on historical first and second adjustment data and corresponding updated equipment status values. The Kalman filter is a recursive algorithm that can estimate the system status from noisy observations. In this application, data is processed through prediction and update steps, and updated equipment status values ​​are calculated based on the observed inputs.

[0082] The updated values ​​of the equipment allocation states are calculated using the aforementioned state update model. Based on these updated values, a first scheduling scheme is generated. Specifically, the predicted state is calculated using the equipment allocation states from the previous moment and the first adjustment parameter, and a second adjustment parameter is incorporated into the update step to correct noise, resulting in an updated value. This updated value represents the allocation ratio of equipment such as bucket wheel reclaimers and belt conveyors. Especially in high-dynamic scenarios, if the uncertainty index of arrival time exceeds a preset threshold, the model will prioritize allocating more equipment to the high-priority sequence, forming a preliminary first scheduling scheme. The first scheduling scheme is then used for target optimization to obtain a second scheduling scheme. The objective optimization includes minimizing loading delay, which is achieved by applying the Particle Swarm Optimization (PSO) algorithm. PSO is an optimization algorithm that simulates the foraging behavior of flocks of birds in nature. It finds the optimal solution by simulating the iterative updates of the position and velocity of individuals in the population within the search space. In the optimization process, the objective function is defined as minimizing loading delay. The optimization process involves initializing the particle swarm, with each particle representing a combination of loading equipment allocations. Then, the fitness of each particle is calculated, which is the delay time calculated based on the scheme. The optimal position of the particles and the global optimal position are updated until the algorithm converges, resulting in a second scheduling scheme. The above optimization method can effectively reduce loading delay and improve resource utilization.

[0083] In this optimization process, if low-dynamic scenarios are involved, particle swarm optimization will incorporate crossover operations of genetic algorithms to increase search diversity and further refine the second scheduling scheme, bringing the delay close to zero. At the same time, if the parameter combination cannot meet the requirements, the system will iteratively incorporate dynamic fusion factors and ensure that the optimization objective remains consistent, further improving the effect of minimizing loading delay. Finally, the second scheduling scheme after objective optimization will be output as the final multi-machine collaborative scheduling scheme. The above-mentioned second scheduling scheme is implemented through the loading equipment control system, ensuring efficient scheduling and resource optimization in a real-time dynamic environment. The above-mentioned second scheduling scheme can achieve real-time and accurate equipment allocation and collaborative scheduling in complex railway loading scenarios based on the fluctuation of train arrival time and equipment operating status, effectively reducing loading delay and improving the overall efficiency of railway loading and unloading operations.

[0084] This invention also provides a multi-machine collaborative automated operation system for train loading, used to implement the above-mentioned method, such as... Figure 4 As shown, the system includes:

[0085] The data acquisition unit is used to acquire real-time location data and estimated arrival time fluctuation information of trains on railway tracks through sensor networks, and to process the real-time location data and estimated arrival time fluctuation information using a time series prediction model to obtain a quantitative index of arrival time uncertainty.

[0086] The priority judgment unit is used to determine the relationship between the arrival time uncertainty quantification index and the preset threshold, and to determine the allocation priority sequence of the loading equipment;

[0087] The scheme optimization unit is used to acquire the allocation priority sequence and the current loading equipment status data, adjust the equipment operating parameters using an optimization algorithm to obtain a preliminary adjustment scheme; and verify whether the preliminary adjustment scheme meets the preset requirements to obtain the optimization parameters.

[0088] The model building unit is used to construct a resource utilization model based on the optimization parameters and multi-machine collaboration requirements to obtain a draft multi-machine collaboration rotation schedule; based on the draft rotation schedule, it determines the relationship between the decision time of the draft and the preset time limit to obtain the final refined schedule;

[0089] The scheduling scheme update unit is used to update the collaboration time between devices based on the final refining schedule and real-time optimization feedback, so as to obtain a multi-machine collaborative scheduling scheme.

[0090] The present invention also provides a computer-readable storage medium storing instructions that, when executed by a processor, implement the above-described method.

[0091] In summary, this invention acquires real-time train location information and estimated arrival time fluctuation data through a sensor network. Using a time series prediction model, it obtains a quantitative index of train arrival time uncertainty. Based on this quantitative index, it compares it with a preset threshold to determine if the current scenario is a high-dynamic scenario, and generates a priority sequence for loading equipment allocation. In high-dynamic scenarios, equipment is prioritized for allocation to track sections with larger fluctuations to cope with sudden changes in arrival time. Based on the priority sequence and equipment status data, an optimization algorithm adjusts the equipment's operating parameters to obtain a preliminary equipment parameter adjustment plan. After verification, this plan ensures that the preset requirements for loading operations are met, thereby generating optimized parameters. This allows the equipment to adjust its operating status promptly and accurately in dynamic scenarios to adapt to constantly changing operational needs. Based on multi-machine collaboration... The system also requires the construction of a resource utilization model to generate a draft multi-machine collaborative rotation schedule, ensuring that the rotation interval matches the loading operation time requirements. If the decision-making time exceeds the preset limit, parallel computing methods are employed to further improve decision-making efficiency. Ultimately, the refined schedule provides a clear time arrangement for subsequent equipment collaborative operations. Based on the final refined schedule and real-time optimization feedback, the collaboration time between equipment is updated, generating a multi-machine collaborative scheduling scheme. This scheme can adjust the collaborative status of equipment in real time, ensuring efficient equipment collaboration and reducing resource waste and delays caused by equipment idleness or mismatched work. Through the mutual cooperation of the above technical solutions, it ensures that equipment can flexibly adjust and collaborate in dynamic railway loading scenarios, thereby effectively improving operational efficiency, reducing delays, and providing technical support for the automation and intelligence of railway loading operations.

[0092] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0093] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0094] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for multi-machine cooperative automated operation of a train loading, characterized in that, The method comprises: Step S1: obtaining real-time position data and predicted arrival time fluctuation information of a train on a railway track through a sensor network, processing the real-time position data and predicted arrival time fluctuation information by using a time series prediction model to obtain an arrival time uncertainty quantification index; Step S2: determining the relationship between the arrival time uncertainty quantification index and a preset threshold value to determine a distribution priority sequence of loading equipment; Step S3: obtaining the distribution priority sequence and current loading equipment state data, adjusting equipment working parameters by using an optimization algorithm to obtain a preliminary adjustment scheme, and verifying whether the preliminary adjustment scheme meets a preset requirement according to the preliminary adjustment scheme to obtain optimization parameters; Step S4: constructing a resource utilization model by using the optimization parameters and multi-machine cooperation requirements to obtain a multi-machine cooperation rotation schedule draft, and determining the relationship between the decision-making time of the rotation schedule draft and a preset time limit to obtain a final refined schedule; Step S5: updating the cooperation time between devices according to the final refined schedule and real-time optimization feedback to obtain a multi-machine cooperation scheduling scheme.

2. The method of claim 1, wherein, Step S1 comprises: Collecting real-time position data and predicted arrival time fluctuation information of a train on a railway track through a sensor network, preprocessing the real-time position data to obtain a first position data set, extracting features from the predicted arrival time fluctuation information to obtain a time fluctuation feature set, inputting the first position data set and the time fluctuation feature set into a pre-trained time series prediction model to obtain a first prediction result, calculating an arrival time uncertainty quantification index according to the first prediction result, wherein the quantification index represents the fluctuation range of the train arrival time, and outputting the arrival time uncertainty quantification index.

3. The method of claim 1, wherein, Step S2 comprises: Determining whether the quantification index is greater than a preset threshold value, if yes, marking it as a high dynamic scenario, otherwise, marking it as a low dynamic scenario, determining a first priority sequence according to the high dynamic scenario or the low dynamic scenario, wherein the first priority sequence includes the sequence of device distribution, sorting the first priority sequence to obtain a second priority sequence, and outputting the second priority sequence as the distribution priority sequence of the device.

4. The method of claim 1, wherein, In step S3, obtaining a preliminary adjustment scheme comprises: Obtaining the distribution priority sequence and current loading equipment state data, inputting the distribution priority sequence and the current loading equipment state data into a pre-established genetic algorithm model, calculating the pitch angle parameter of the bucket wheel reclaimer and the speed parameter of the belt conveyor by using the genetic algorithm model to obtain a first parameter combination, and the genetic algorithm model uses a real number coding genetic algorithm; Evaluating the fitness of the first parameter combination to obtain a first fitness value, and iteratively adjusting the parameters in the genetic algorithm model according to the first fitness value to obtain a second parameter combination as the preliminary adjustment scheme.

5. The method of claim 4, wherein, In step S3, verifying whether the preliminary adjustment scheme meets a preset requirement according to the preliminary adjustment scheme to obtain optimization parameters comprises: Extract a first parameter combination from the preliminary adjustment scheme; obtain a preset emergency demand response attribute; determine whether the first parameter combination meets the emergency demand response attribute, if yes, determine the first parameter combination as an optimization parameter; if not, iteratively adjust the genetic algorithm model, incorporate a dynamic fusion factor, and obtain a second parameter combination; verify the second parameter combination to obtain a third parameter combination, and use the third parameter combination as an optimization parameter.

6. The method of claim 1, wherein, In step S4, a multi-machine cooperative rotation schedule draft is obtained, including: Obtain the optimization parameter and multi-machine cooperation requirements; determine the operating state parameters of each device according to the optimization parameter; determine the cooperative constraint conditions between devices according to the multi-machine cooperation requirements; input the operating state parameters and the cooperative constraint conditions into a pre-established resource utilization model; calculate the rotation interval between devices through the resource utilization model to obtain a first rotation schedule; perform constraint verification on the first rotation schedule to obtain a second rotation schedule; and output the second rotation schedule as a multi-machine cooperative rotation schedule draft.

7. The method of claim 6, wherein, In step S4, a final refined schedule is obtained, including: Obtain the rotation schedule draft; calculate the decision time of the rotation schedule draft; if the decision time is greater than a preset time limit, use a parallel computing method to speed up the parameter calculation process to obtain a first refined schedule; if the decision time is less than or equal to the preset time limit, determine the rotation schedule draft as a first refined schedule; perform constraint optimization on the first refined schedule to obtain a second refined schedule; and output the second refined schedule as a final refined schedule.

8. The method of claim 1, wherein, In step S5, a multi-machine cooperative scheduling scheme is obtained, including: Obtain the final refined schedule and real-time optimization feedback data; determine a first adjustment parameter of the device allocation state according to the real-time optimization feedback data; calculate a second adjustment parameter of the device allocation state according to the final refined schedule; input the first adjustment parameter and the second adjustment parameter into a pre-established state update model; calculate the update value of the device allocation state through the state update model to obtain a first scheduling scheme; perform target optimization on the first scheduling scheme to obtain a second scheduling scheme, wherein the target optimization includes minimizing the loading delay; and use the second scheduling scheme as a multi-machine cooperative scheduling scheme.

9. A multi-machine coordinated automated job system for train loading, for implementing the method of any one of claims 1-8, characterized in that, The system includes: A data acquisition unit configured to acquire real-time position data and predicted arrival time fluctuation information of a train on a railway track through a sensor network, and to process the real-time position data and the predicted arrival time fluctuation information using a time series prediction model to obtain an arrival time uncertainty quantization index; A priority judgment unit configured to determine a relationship between the quantization index and a preset threshold based on the arrival time uncertainty quantization index, and to determine a sequence of allocation priorities of loading devices. The scheme optimization unit is configured to obtain the allocation priority sequence and current loading equipment state data, adjust equipment working parameters by using an optimization algorithm to obtain a preliminary adjustment scheme, and check whether the preliminary adjustment scheme meets preset requirements according to the preliminary adjustment scheme to obtain optimization parameters. The model construction unit is configured to construct a resource utilization model by using the optimization parameters and multi-machine coordination requirements to obtain a multi-machine coordination rotation schedule draft, and determine a relationship between decision-making time of the rotation schedule draft and preset time limits to obtain a final refined schedule. The scheduling scheme updating unit is configured to update cooperation time between equipment according to the final refined schedule and real-time optimization feedback to obtain a multi-machine coordination scheduling scheme.

10. A computer-readable storage medium having stored thereon instructions, the instructions comprising, The instructions are executed by the processor to implement the method in any one of claims 1-8.

Images