CN118246711B

CN118246711B - Intelligent scheduling method and intelligent scheduling system for construction equipment

Info

Publication number: CN118246711B
Application number: CN202410669310.6A
Authority: CN
Inventors: 刘魁; 任杰; 孙帅; 梅斌
Original assignee: Sany Automobile Manufacturing Co Ltd
Current assignee: Sany Automobile Manufacturing Co Ltd
Priority date: 2024-05-28
Filing date: 2024-05-28
Publication date: 2024-08-06
Anticipated expiration: 2044-05-28
Also published as: CN118246711A

Abstract

The application provides an intelligent scheduling method and an intelligent scheduling system for construction equipment, which solve or improve the technical problem that the scheduling of mine construction equipment in the prior art lacks a prediction mode for future conditions. The intelligent scheduling method for the construction equipment integrates time sequence characteristics of different scales, and can more accurately and stably predict the time sequence of each target construction equipment. And inputting the predicted time sequence data of each target construction equipment into an operation optimization model based on the corresponding constraint conditions to perform operation calculation so as to determine the scheduling strategies of a plurality of target construction equipment, and generating the control instruction of each target construction equipment according to the scheduling strategies so that the target construction equipment works with the corresponding control instruction. The method has the advantages that through the combination of prediction and model optimization, the working requirements can be predicted in advance, the scheduling strategy is optimized and adjusted, and the working efficiency and the resource utilization rate of the mine working equipment are improved.

Description

Intelligent scheduling method and intelligent scheduling system for construction equipment

Technical Field

The application relates to the technical field of engineering construction, in particular to an intelligent scheduling method and an intelligent scheduling system for construction equipment.

Background

At present, when the surface mine is exploited and the earthwork is constructed, the transport vehicle needs to be charged beside the excavator, and the transport vehicle needs to be discharged to a corresponding discharging area after the charging is completed. Different types of loading have different discharge areas, such as ore to be discharged to the ore zone and clinker to be discharged to the slag zone. Therefore, the scheduling of mine construction equipment, which is the scheduling of individual construction equipment (e.g., transport vehicles, etc.), is a critical task in mine engineering, which involves the efficient organization, coordination, and optimization of various kinds of equipment to ensure the smooth progress of the construction process.

Currently, scheduling of mine construction equipment (e.g., excavators, transport vehicles) and the like is often performed empirically by hand, is inefficient, and is prone to error.

In order to improve efficiency, a machine algorithm system and other auxiliary personnel are adopted to schedule the equipment, but most of machine algorithms utilize monitoring data of the current working condition, and simple judgment logic rules are adopted to schedule the construction equipment, but the equipment is limited by human factors in actual application and has poor floor property. And a prediction mode for future conditions is lacking, so that a globally optimal scheduling scheme is difficult to obtain.

Disclosure of Invention

In view of the above, the application provides an intelligent scheduling method and an intelligent scheduling system for construction equipment, which solve or improve the technical problems that in the prior art, the scheduling of mine construction equipment is limited by artificial factors, the floor is poor, a prediction mode for future conditions is lacking, and a globally optimal scheduling scheme is difficult to obtain.

As a first aspect of the present application, the present application provides an intelligent scheduling method for construction equipment, including:

Invoking historical operation data of target construction equipment and historical operation data of at least one first construction equipment of the same equipment as the target construction equipment, wherein the historical operation data comprises a plurality of historical preset time lengths and historical time sequence data of the construction equipment within the historical preset time lengths, and the historical time sequence data comprises time points arranged according to time sequences and working condition data corresponding to the time points; fusing the historical operation data of the target construction equipment with the historical operation data of at least one first construction equipment to generate fused time sequence data; fusing the historical operation data of the target construction equipment and the fused time sequence data to generate predicted time sequence data of the target construction equipment within a target preset duration; acquiring constraint conditions corresponding to a plurality of target construction devices in a preset scheduling target time period, and calling the operation optimization solver to perform simulation calculation on predicted time sequence data of the plurality of target construction devices in the preset target time period based on the constraint conditions so as to output scheduling strategies of the plurality of target construction devices in the preset target time period; and generating control instructions of each target construction equipment based on the scheduling policy, so that each target construction equipment executes corresponding tasks with the corresponding control instructions.

In one possible implementation manner of the present application, fusing the historical operation data of the target construction device with the historical operation data of at least one first construction device to generate fused time series data includes:

Inputting the historical operation data of the target construction equipment and the historical operation data of at least one first construction equipment into a convolutional neural network for convolutional calculation, and outputting first data; inputting the first data into a first GRU network for processing, and outputting first predicted time sequence data of the target construction equipment within a target preset duration; inputting the historical operation data of the target construction equipment into an autoregressive model for prediction, and outputting second prediction time sequence data of the target construction equipment within a target preset duration; and inputting the first predicted time sequence data and the second predicted time sequence data into a first full-connection layer for fusion, and outputting fusion time sequence data.

Inputting the historical operation data of the target construction equipment and the historical operation data of at least one first construction equipment into a convolutional neural network for convolutional calculation, and outputting first data; inputting the first data into a first GRU network for processing, and outputting first predicted time sequence data of the target construction equipment within a target preset duration; inputting the first predicted time sequence data into a second GRU network for processing, and outputting third predicted time sequence data, wherein the second GRU network is a cycle skip GRU network; inputting the historical operation data of the target construction equipment into an autoregressive model for prediction, and outputting second prediction time sequence data of the target construction equipment within a target preset duration; and inputting the first predicted time sequence data, the second predicted time sequence data and the third predicted time sequence data into a first full-connection layer for fusion, and outputting fusion time sequence data.

inputting the historical operation data of the target construction equipment and the historical operation data of at least one first construction equipment into a convolutional neural network for convolutional calculation, and outputting first data; inputting the first data into a first GRU network for processing, and outputting first predicted time sequence data of the target construction equipment within a target preset duration; inputting the first predicted time sequence data into a second GRU network for processing, and outputting third predicted time sequence data, wherein the second GRU network is a cycle skip GRU; and inputting the first predicted time sequence data and the third predicted time sequence data into a first full-connection layer for fusion, and outputting fusion time sequence data.

In one possible implementation manner of the present application, the historical job data further includes historical space-time sequence data, where the historical space-time sequence data includes time points arranged according to a time sequence and space positions corresponding to the time points, and the intelligent scheduling method further includes: inputting the historical space time series data of the target construction equipment and the historical space time series data of at least one first construction equipment into a space prediction model for prediction to generate predicted space time series data; and fusing the historical operation data, the fusion time series data and the prediction space time series data of the target construction equipment to generate prediction time series data of the target construction equipment within a target preset duration.

In one possible implementation manner of the present application, after generating the control instruction of each target construction device based on the scheduling policy, so that each target construction device performs a corresponding task with the corresponding control instruction, the intelligent scheduling method further includes: real-time sequence data of target construction equipment in a target preset duration are detected in real time; and when the real-time sequence data of the target construction equipment in the target preset time length is different from the predicted time sequence data of the target construction equipment, calling the operation optimization solver to adjust a scheduling strategy according to the real-time sequence data of the target construction equipment.

As a second aspect of the present application, the present application also provides an intelligent scheduling system for construction equipment, including:

The database is used for storing the operation data of the construction equipment; the model library comprises a first submodule, wherein the first submodule is used for calling historical operation data of target construction equipment; the second sub-module is used for calling historical operation data of at least one first construction device of the same type of equipment as the target construction device; the third submodule is used for fusing the historical operation data of the target construction equipment with the historical operation data of at least one first construction equipment to generate fused time sequence data; the fourth submodule is used for fusing the historical operation data of the target construction equipment with the fused time sequence data to generate predicted time sequence data of the target construction equipment within a target preset duration; and an operational optimization model, the operational optimization model comprising: the system comprises a constraint condition module, a day-ahead scheduling module and an operation planning optimization solver, wherein the constraint condition module is used for selecting constraint conditions; the day-ahead scheduling module is used for acquiring constraint conditions corresponding to a plurality of target construction devices within a preset scheduling target time length, calling the operation optimization solver to perform simulation calculation on predicted time sequence data of the plurality of target construction devices within the preset target time length based on the constraint conditions so as to output a scheduling strategy of the plurality of target construction devices within the preset target time length, and generating a control instruction of each target construction device based on the scheduling strategy so that each target construction device executes a corresponding task with the corresponding control instruction.

In one possible implementation manner of the present application, the operation optimization model further includes: the real-time scheduling model is used for detecting real-time sequence data of the target construction equipment in a target preset duration in real time; and when the real-time sequence data of the target construction equipment in the target preset time length is different from the predicted time sequence data of the target construction equipment, calling the operation optimization solver to adjust a scheduling strategy according to the real-time sequence data of the target construction equipment.

In one possible implementation manner of the present application, the historical job data further includes historical space time series data, where the historical space time series data includes time points arranged according to a time series and space positions corresponding to the time points; wherein the model library further comprises: the spatial prediction model is used for predicting the historical spatial time series data of the target construction equipment and the historical spatial time series data of at least one first construction equipment to generate predicted spatial time series data; the fourth submodule is used for fusing the historical operation data of the target construction equipment, the fusion time series data and the prediction space time series data to generate prediction time series data of the target construction equipment within a target preset duration.

In one possible implementation of the present application, the third submodule includes: the convolution neural network is used for carrying out convolution calculation on the historical operation data of the target construction equipment and the historical operation data of at least one first construction equipment and outputting first data; the first GRU network is used for processing the first data input and outputting first predicted time sequence data of the target construction equipment within a target preset duration; and/or a second GRU network for processing the first predicted time-series data and outputting third predicted time-series data, wherein the second GRU network is a cycle-skip GRU; and/or an autoregressive model, which is used for predicting the historical operation data of the target construction equipment and outputting second predicted time sequence data of the target construction equipment within a target preset duration; and the first full-connection layer is used for fusing the first predicted time sequence data, the second predicted time sequence data and/or the third predicted time sequence data and outputting fused time sequence data.

According to the intelligent scheduling method for the construction equipment, the target construction equipment and the historical operation data of the same type of construction equipment are adopted, the target construction equipment and the historical operation data of the same type of construction equipment are fused, the fused time series data and the historical operation data (namely, the historical time series data) of the target construction equipment are fused again, and the predicted time series data, namely, the predicted time series data of the target construction equipment, are determined, so that data with higher relativity and periodicity are obtained, namely, the time series characteristics of different scales are fused, and the time series prediction of each target construction equipment can be performed more accurately and stably. And finally, inputting the predicted time sequence data of each target construction device into an operation optimization model based on the corresponding constraint conditions to perform operation calculation so as to determine the scheduling strategies of a plurality of target construction devices, and generating the control instruction of each target construction device according to the scheduling strategies so that the target construction device operates with the corresponding control instruction. The method has the advantages that through the combination of prediction and model optimization, the working requirements can be predicted in advance, the scheduling strategy is optimized and adjusted, and the working efficiency and the resource utilization rate of the mine working equipment are improved.

Drawings

The above and other objects, features and advantages of the present application will become more apparent by describing embodiments of the present application in more detail with reference to the attached drawings. The accompanying drawings are included to provide a further understanding of embodiments of the application and are incorporated in and constitute a part of this specification, illustrate the application and together with the embodiments of the application, and not constitute a limitation to the application. In the drawings, like reference numerals generally refer to like parts or steps.

Fig. 1 is a schematic flow chart of an intelligent scheduling method for construction equipment according to an embodiment of the present application.

Fig. 2 is a flow chart of an intelligent scheduling method for construction equipment according to another embodiment of the present application.

Fig. 3 is a flow chart of an intelligent scheduling method for construction equipment according to another embodiment of the present application.

Fig. 4 is a flow chart of an intelligent scheduling method for construction equipment according to another embodiment of the present application.

Fig. 5 is a flow chart of an intelligent scheduling method for construction equipment according to another embodiment of the present application.

Fig. 6 is a flow chart of an intelligent scheduling method for construction equipment according to another embodiment of the present application.

Fig. 7 is a schematic diagram of the operation of a third sub-module in the dispatching system of construction equipment according to an embodiment of the present application.

Fig. 8 is a schematic diagram of the operation of a third sub-module in a model library in a dispatching system of construction equipment according to another embodiment of the present application.

Fig. 9 is a schematic diagram of operation of a model library in a dispatching system of construction equipment according to another embodiment of the present application.

Fig. 10 is a schematic diagram of operation of a model library in a dispatching system for construction equipment according to another embodiment of the present application.

Fig. 11 is a schematic diagram of operation of a model library in a dispatching system for construction equipment according to another embodiment of the present application.

Fig. 12 is a schematic diagram of operation of an operational optimization model in a dispatching system of construction equipment according to another embodiment of the present application.

Fig. 13 is a schematic diagram of an electronic device according to an embodiment of the present application.

Detailed Description

In the description of the present application, the meaning of "plurality" means at least two, for example, two, three, etc., unless specifically defined otherwise. Furthermore, the terms "comprise" and "have," as well as any variations thereof, are intended to cover a non-exclusive inclusion. For example, a process, method, system, article, or apparatus that comprises a list of steps or elements is not limited to only those listed steps or elements but may include other steps or elements not listed or inherent to such process, method, article, or apparatus.

Furthermore, references herein to "an embodiment" mean that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the present application. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. Those of skill in the art will explicitly and implicitly appreciate that the embodiments described herein may be combined with other embodiments.

Exemplary method

As a first aspect of the present application, the present application provides an intelligent scheduling method for construction equipment, fig. 1 is a schematic flow chart of an intelligent scheduling method for construction equipment, as shown in fig. 1, and the intelligent scheduling method for construction equipment includes the following steps:

S1: and calling the historical operation data of the target construction equipment and the historical operation data of at least one first construction equipment of the same equipment as the target construction equipment.

The historical job data comprise a plurality of historical preset time lengths and historical time sequence data of target construction equipment within the historical preset time lengths, wherein the historical time sequence data comprise time points arranged according to time sequences and working condition data corresponding to the time points. Specifically, the preset time period may be one day, and then the plurality of preset time periods may be 10 days, etc. For example, the historical job data includes time-series data of the target construction equipment for each day within 10 days of the history.

Specifically, the historical job data refers to job data before the current predicted time point, for example, if the predicted time series data of the target construction equipment on the day is predicted after the end of the work today, the historical job data is the job data of the target construction equipment on the day and the current previous job data is the historical job data.

Specifically, the target construction equipment is one of a plurality of construction equipment under one working condition of mine construction, such as an excavator.

The first construction equipment is the same kind of construction equipment as the target construction equipment under the working condition. For example, the target construction equipment is a first excavator, and then the first construction equipment is a second excavator, a third excavator and the like under the same working condition.

Specifically, the historical job data generally refers to data of the target construction equipment before the current time, for example, if predicted time series data of the target construction equipment of 2023, 10, 16 days is predicted, then the historical job data is the historical job data of N days (for example, the first 10 days) before 2023, 10, 16 days.

The time series data refers to time points arranged according to time series and working condition data corresponding to the time points, for example, the time series data of the excavator at 2023, 10 and 15 days includes: at each time point of 2023, 10, 15 and at each time point, the working condition data (e.g. excavation yield) of the excavator, and for example, the time series data of 2023, 10, 15 of the unloader include: at each time point of day 10, 15 of 2023 and at each time point, the operating conditions of the unloader (e.g., the unloading output).

Specifically, working condition data (such as a position, a load, a working state, a working speed, a device health state and the like of each construction device under one working condition) of each construction device are detected through corresponding vehicle-mounted sensors, the vehicle-mounted sensors detect the working condition data of the construction devices in real time, and the working condition data are transmitted to a database in a remote server through data transmission equipment and stored. When the historical job data of the target construction equipment and the historical job data of the first construction equipment are scheduled, the historical job data of the target construction equipment and the historical job data of the first construction equipment can be directly called in the database.

S2: fusing the historical operation data of the target construction equipment with the historical operation data of at least one first construction equipment to generate fused time sequence data;

after the historical operation data of the target construction equipment and the historical operation data of the first construction equipment are fused, fusion time series data can be generated, namely, the prediction time series data of the target construction equipment can be predicted by synchronously referring to the historical operation data of the same type of construction equipment.

S3: fusing the historical operation data and the fusion time series data of the target construction equipment to generate predicted time series data of the target construction equipment within a target preset duration;

specifically, the target preset duration may be one day, for example, after the work on the same day is finished, the historical operation data of the same day and the historical operation data before the same day are used to predict the predicted time sequence data of the target construction equipment in the open day, that is, predict the open day operation condition of the target construction equipment.

And re-fusing the historical operation data and the fused time series data of the target construction equipment to generate predicted time series data, namely, the operation data of the target construction equipment on the open day is predicted to be completed.

And finishing the predicted time series data of other construction equipment in the working condition according to the modes S1-S3.

After the predicted time series data of each construction equipment in the working condition are completed, overall calculation can be performed according to the predicted time series data of each construction equipment so as to determine the scheduling mode of each construction equipment in the working condition, namely S4-S5 are executed.

S4: acquiring constraint conditions corresponding to a plurality of target construction devices in a target preset time length, and calling an operation optimization solver based on the constraint conditions to perform simulation calculation on predicted time sequence data of the plurality of target construction devices in the target preset time length so as to output a scheduling strategy of the plurality of target construction devices in the target preset time length;

specifically, the constraint condition refers to: under the working condition, a plurality of construction equipment schedule needs the condition that satisfies. Constraints may include:

(1) General constraints: refers to general constraint conditions under different working conditions, such as capacity constraint, time constraint, demand constraint and the like.

(2) Customizing constraint conditions: refers to specific constraint conditions under special working conditions and constraint conditions customized according to user requirements. For example: the constraint of reducing electricity cost for peak-shifting charging of electric equipment is overcome.

(3) Uncertainty factor: refers to a constraint condition specified by an uncertain factor such as equipment failure, for example: constraints formulated for equipment downtime, constraints formulated for order changes, constraints formulated for blackouts.

And inputting constraint conditions and predicted time sequence data of each construction device within a target preset duration into an operation optimizing model, and calling a solver to carry out simulation calculation solution so as to output scheduling strategies of a plurality of construction devices.

S5: based on the scheduling policy, control instructions for each target construction equipment are generated such that each target construction equipment performs a corresponding task with the corresponding control instructions.

After the scheduling strategy is finished, a control instruction of each target construction equipment can be generated according to the scheduling strategy, then the control instruction is transmitted to a controller of the target construction equipment through the data transmission equipment, and the controller controls the construction equipment to work according to the control instruction.

In one possible implementation manner of the present application, the specific process of fusing the historical operation data of the target construction equipment with the historical operation data of the first construction equipment of the same kind may adopt the following three modes:

(1) As shown in fig. 2 and 8, S2 (fusing the historical job data of the target construction equipment with the historical job data of at least one first construction equipment to generate fused time-series data) specifically includes the steps of:

s21: inputting the historical operation data of the target construction equipment and the historical operation data of at least one first construction equipment into a convolutional neural network (namely a CNN network) for convolutional calculation, and outputting first data;

Namely, the historical time series data in the historical job data of the target construction equipment and the historical time series data in the historical job data of at least one first construction equipment (of the same kind as the target construction equipment) include: for example, the historical time series data of YYYY year, M month, D, d+1 day of the first excavator (target construction equipment), and the historical time series data of … … yyyyyy year, M month, d+n days of the first excavator (target construction equipment). Historical time series data of yyyyy year, M month, D, and historical time series data of yyyyy year, M month, d+1, and historical time series data of … … yyyyyy year, M month, d+n for the second excavator (first construction equipment). Historical time series data of yyyyy year, M month, D, and historical time series data of yyyyy year, M month, d+1, and historical time series data of … … yyyyyy year, M month, d+n for the third excavator (first construction equipment). That is, the historical job data set of the target construction equipment within the history n+1 days (i.e., the historical time series data) is: (X ¹,X²……Xⁿ⁺¹).

The historical job data set of the first construction equipment within the history n+1 days (i.e., the historical time series data) is: (L ₁,L₂……L_n+1).

The method comprises the steps of inputting a plurality of historical time series data sets of target construction equipment and a plurality of historical time series data sets of at least one first construction equipment into a convolutional neural network (namely a CNN network) for convolutional calculation, and outputting first data. The first data is still the fused time series data.

S22: inputting the first data into a first GRU network for processing, and outputting first predicted time sequence data of target construction equipment within target preset duration;

Specifically, the first GRU network is a gated loop unit (Gated Recurrent Unit) network, which is a special type of loop neural network (RNN). In contrast to conventional RNNs, the GRU network controls the flow of information by introducing two gating units, an Update Gate and a reset Gate (RESET GATE).

The GRU network has strong nonlinear mapping capability, can handle complex nonlinear problems, and has wide application in various fields, such as time sequence prediction, natural language processing, voice recognition and the like.

The GRU network can predict future values from past observations.

When the fused first data is input to the first GRU network, the first GRU network captures long-term dependency relationship in the time sequence in the first data through the gating unit, and outputs a final prediction result, namely, outputs first prediction time sequence data.

S23: inputting historical operation data of the target construction equipment into an autoregressive model for prediction, and outputting second prediction time sequence data of the target construction equipment within a target preset duration;

An autoregressive model (Autoregressive Model, abbreviated as AR model) is a time series analysis method that uses past data of the variable itself to predict its future value.

The basic principle of the autoregressive model is as follows: each data point in the time series has a linear relationship with its preceding data point. By identifying these linear relationships and using them, trends and periodicity in the sequence can be captured, and predictions made.

Therefore, a historical time series data set (X ¹,X²……Xⁿ⁺¹) of the target construction equipment in the history n+1 days is input into the autoregressive model, the historical time series data set of the target construction equipment is taken as input, the linear relation among each historical time series data set is identified, the trend and the periodicity in the historical time series data set are captured by utilizing the linear relation, and further second predicted time series data of the target construction equipment in the target preset time period are predicted.

S24: and inputting the first predicted time sequence data and the second predicted time sequence data into the first full-connection layer for fusion, and outputting the fusion time sequence data.

The full connection layer converts the feature images or feature vectors output by the first GRU network and the CNN network into one-dimensional output vectors, so that the prediction of the time sequence is realized.

The first predicted time-series data is predicted by the first GRU network for the fused first data.

The second predicted time-series data is predicted by the autoregressive model according to the historical time-series data set of the target construction equipment.

The first full connection layer fuses the first predicted time series data and the second predicted time series data and outputs final fused time series data.

(2) As shown in fig. 3 and 9, S2 (fusing the historical job data of the target construction equipment with the historical job data of at least one first construction equipment to generate fused time-series data) specifically includes the steps of:

S210: inputting the historical operation data of the target construction equipment and the historical operation data of at least one first construction equipment into a convolutional neural network for convolutional calculation, and outputting first data;

S220: inputting the first data into a first GRU network for processing, and outputting first predicted time sequence data of target construction equipment within target preset duration;

In the same way as S22, when the fused first data is input to the first GRU network, the first GRU network captures the long-term dependency relationship in the time sequence in the first data through the gate control unit, and outputs the final prediction result, that is, outputs the first predicted time sequence data.

S230: inputting the first predicted time sequence data into a second GRU network for processing, and outputting third predicted time sequence data, wherein the second GRU network is a cycle skip GRU network;

the second GRU network is a periodically-hopped GRU network, which is a GRU network having a structure different from that of the first GRU network, and the periodically-hopped GRU network (Periodic Skip-connection GRU Network) is a GRU network structure combining periodicity and hopping connections (Skip-connections) for better processing time-series data having a Periodic pattern. For example, the time series of monday is correlated with yesterday, i.e., sunday, but may be more correlated with monday. The second GRU is more likely to capture such periodic trends.

And inputting the first predicted time series data obtained through the first GRU network processing into a second GRU network for processing, and outputting third predicted time series data. Because the second GRU network is a periodic jump GRU network, the time sequence data with obvious periodicity can be predicted more accurately, and the accuracy of the predicted time sequence data of the target construction equipment is further improved.

S240: inputting historical operation data of the target construction equipment into an autoregressive model for prediction, and outputting second prediction time sequence data of the target construction equipment within a target preset duration;

And S23, inputting a historical time series data set (X ¹,X²……Xⁿ⁺¹) of the target construction equipment in a history n+1 days into an autoregressive model, taking the historical time series data set of the target construction equipment as input, identifying a linear relation among each historical time series data set, capturing a trend and periodicity in the historical time series data set by utilizing the linear relation, and further predicting second predicted time series data of the target construction equipment in a target preset duration.

S250: and inputting the first predicted time sequence data, the second predicted time sequence data and the third predicted time sequence data into the first fully-connected layer for fusion, and outputting the fusion time sequence data.

And S24, the full connection layer converts the feature images or feature vectors output by the first GRU network and the CNN network into one-dimensional output vectors, so that the prediction of the time sequence is realized. Therefore, the first full connection layer fuses the first predicted time-series data, the second predicted time-series data, and the third predicted time-series data and outputs final fused time-series data.

(3) As shown in fig. 4 and 10, S2 (fusing the historical job data of the target construction equipment with the historical job data of at least one first construction equipment to generate fused time-series data) specifically includes the steps of:

s25: inputting the historical operation data of the target construction equipment and the historical operation data of at least one first construction equipment into a convolutional neural network for convolutional calculation, and outputting first data;

S26: inputting the first data into a first GRU network for processing, and outputting first predicted time sequence data of target construction equipment within target preset duration;

in the same way as in S22 above, when the fused first data is input to the first GRU network, the first GRU network captures the long-term dependency relationship in the time sequence in the first data through the gate control unit, and outputs the final prediction result, that is, outputs the first predicted time sequence data.

S27: inputting the first predicted time sequence data into a second GRU network for processing, and outputting third predicted time sequence data, wherein the second GRU network is a cycle skip GRU;

In the same manner as in S230, the first predicted time-series data obtained by the first GRU network processing is inputted into the second GRU network and processed therein, and the third predicted time-series data is outputted. Because the second GRU network is a periodic jump GRU network, the time sequence data with obvious periodicity can be predicted more accurately, and the accuracy of the predicted time sequence data of the target construction equipment is further improved.

S28: and inputting the first predicted time sequence data and the third predicted time sequence data into the first full-connection layer for fusion, and outputting the fusion time sequence data.

And S24, the full connection layer converts the feature images or feature vectors output by the first GRU network and the CNN network into one-dimensional output vectors, so that the prediction of the time sequence is realized. Therefore, the first full connection layer fuses the first predicted time-series data and the third predicted time-series data and outputs final fused time-series data.

In one possible implementation manner of the present application, as shown in fig. 5, the historical job data further includes historical space-time sequence data, where the historical space-time sequence data includes time points arranged according to a time sequence and corresponding space positions at the time points, and the intelligent scheduling method further includes the following steps:

S6: inputting the historical space time series data of the target construction equipment and the historical space time series data of at least one first construction equipment into a space prediction model for prediction to generate predicted space time series data;

specifically, the spatial prediction model can predict distribution and change of the spatial pattern, that is, predict the predicted spatial time series data of the target construction equipment within the target preset duration, based on the historical spatial time series data of the target construction equipment and the historical spatial time series data of the first construction equipment.

S7: and fusing the historical operation data, the fusion time series data and the prediction space time series data of the target construction equipment to generate the prediction time series data of the target construction equipment within the target preset duration.

The method comprises the steps of adopting a first full-connection layer to fuse historical operation data (comprising historical time sequence data and historical space time sequence data) of target construction equipment, fused time sequence data obtained in the step S2 and predicted space time sequence data obtained in the step S6, and obtaining predicted time sequence data of the target construction equipment within target preset duration.

By adopting the spatial prediction model, when the time sequence data is predicted, the time sequence prediction can be more accurately and stably performed by fusing time sequence features and spatial features of different scales.

And after the predicted time sequence data in the target preset duration is obtained in the step S7, overall calculation can be carried out according to the predicted time sequence data of each construction device so as to determine the scheduling mode of each construction device in the working condition, namely, the steps S4-S5 are executed.

In one possible implementation manner of the present application, as shown in fig. 6, after S5 (based on the scheduling policy, the control instruction of each target construction device is generated so that each target construction device performs a corresponding task with the corresponding control instruction), the intelligent scheduling method further includes the following steps:

s8: real-time sequence data of target construction equipment in a target preset duration are detected in real time;

and when each construction equipment works to execute the task by the corresponding control instruction within the target preset time length, detecting real-time sequence data of the target construction equipment in real time.

S9: and when the real-time sequence data of the target construction equipment in the target preset time length is different from the predicted time sequence data of the target construction equipment, calling an operation optimization solver to adjust a scheduling strategy according to the real-time sequence data of the target construction equipment.

When the construction equipment works with the predicted time sequence data, real-time sequence data is detected in real time, and a scheduling strategy is timely adjusted according to the real-time sequence data, so that a control instruction of each construction equipment is adjusted, and the scheduling strategy is dynamically adjusted according to the real-time working condition of the construction equipment.

Exemplary System

As a second aspect of the present application, the present application also provides a construction equipment intelligent scheduling system, as shown in fig. 7 to 12, the construction equipment intelligent scheduling system 100 comprising:

A database 1, the database 1 being used for storing job data of construction equipment;

the database 1 may be in communication connection with the vehicle-mounted sensor 400 disposed on each construction device 500, where the vehicle-mounted sensor 400 transmits the detected working condition data of each construction device 500 to the database 1 through the data transmission device 300, and the database 1 stores the working condition data of the construction device after receiving the working condition data of the construction device.

The database 1 may also be communicatively connected to the remote server 200, and the in-vehicle sensor 400 transmits the detected working condition data of each construction equipment to the remote server 200 through the data transmission device 300, where the remote server 200 is communicatively connected to the database 1, and the database 1 invokes the historical working data of the construction equipment in the remote server 200.

The model base 2, the model base 2 includes a first sub-module 21, the first sub-module 21 is used for calling the historical operation data of the target construction equipment; a second sub-module 22, wherein the second sub-module 22 is used for calling historical operation data of at least one first construction equipment of the same type of equipment as the target construction equipment; the third sub-module 23, the third sub-module 23 is configured to fuse the historical operation data of the target construction equipment with the historical operation data of at least one first construction equipment, and generate fused time series data; the fourth sub-module 24 is configured to fuse historical operation data of the target construction device with the fused time sequence data, and generate predicted time sequence data of the target construction device within a target preset duration;

The model library 2 is used for executing S1-S3 in the intelligent scheduling method of the construction equipment, namely the first sub-module 21 executes the historical working condition data of the calling target construction equipment in S1, namely the first sub-module 21 calls the historical working condition data of the target construction equipment from the database 1. The second sub-module 22 executes the historical operating condition data of the at least one first construction equipment (of the same kind as the target construction equipment) called in S1, i.e. the second sub-module 22 calls the historical operating condition data of the at least one first construction equipment from the database 1.

The third sub-module 23 is configured to execute S2, i.e. fuse the historical job data of the target construction equipment with the historical job data of at least one first construction equipment, to generate fused time series data. The fourth sub-module 24 is configured to perform S3, that is, fuse historical job data of the target construction device with the fused time series data, and generate predicted time series data of the target construction device within the target preset duration.

The fourth sub-module 24 is configured to execute S3, i.e. fuse the historical job data of the target construction equipment with the fusion time series data, and generate predicted time series data of the target construction equipment within the target preset duration. In particular, the fourth sub-module 24 may be the second full connection layer 241.

The model library 2 may output predicted time-series data of the target construction equipment.

Operation-specific optimization model 3, the operation-specific optimization model 3 includes: constraint module 31, day-ahead scheduling module 32, and operations optimization solver 33. The constraint condition module 31 is used for selecting constraint conditions; the day-ahead scheduling module 32 is configured to obtain constraint conditions corresponding to a plurality of target construction devices within a preset target duration, and call the operation optimization solver 33 to perform simulation calculation on predicted time sequence data of the plurality of target construction devices within the preset target duration based on the constraint conditions, so as to output predicted time sequences of the plurality of target construction devices within the preset target duration, and generate a control instruction of each target construction device based on the predicted time sequences of each target construction device within the preset target duration, so that each target construction device executes a corresponding task with the corresponding control instruction.

The operation optimization model 3 includes a constraint module 31, a day-ahead scheduling module 32, and an operation optimization solver 33.

The constraint condition module 31 is configured to execute the constraint condition corresponding to the plurality of target construction devices within the target preset duration in S4 in the intelligent scheduling method for construction devices, where constraint condition module 31 stores constraint conditions that need to be met when the construction devices are scheduled, for example, constraint condition module 31 may further include a general constraint condition module 311 configured to select general constraints (refer to general constraints under different working conditions, for example, capacity constraints, time constraints, demand constraints, etc.), a customized constraint module 312 configured to select customized constraints (refer to specific constraints under special working conditions, and constraints customized according to user requirements, for example, constraints that reduce electricity consumption for peak-shifting charging of electric devices), and an uncertainty factor module 313 configured to select uncertainty factors (refer to constraints specified by uncertainty factors such as device faults, etc.), for example, constraints formulated to cope with a machine down, constraints formulated to cope with an order change, constraints formulated to cope with a power outage).

The day-ahead scheduling module 32 is configured to execute the constraint condition-based constraint condition in S4 in the above-described intelligent scheduling method for construction equipment, and call the operation optimization solver 33 to perform simulation calculation on predicted time sequence data of the multiple target construction equipment within the target preset duration, so as to output a scheduling policy of the multiple target construction equipment within the target preset duration.

After the scheduling policy is completed, a control instruction of each target construction device may be generated according to the scheduling policy, and then the control instruction is transmitted to a controller of the target construction device 500 through the data transmission device 300, and the controller controls the construction device 500 to operate according to the control instruction.

According to the intelligent scheduling system for the construction equipment, the target construction equipment and the historical operation data of the same type of construction equipment are adopted, the target construction equipment and the historical operation data of the same type of construction equipment are fused, the fused time series data and the historical operation data (namely, the historical time series data) of the target construction equipment are fused again, and the predicted time series data, namely, the predicted time series data of the target construction equipment, are determined, so that data with higher relativity and periodicity are obtained, namely, the time series characteristics of different scales are fused, and the time series prediction of each target construction equipment can be performed more accurately and stably. And finally, inputting the predicted time sequence data of each target construction device into an operation optimization model based on the corresponding constraint conditions to perform operation calculation so as to determine the scheduling strategies of a plurality of target construction devices, and generating the control instruction of each target construction device according to the scheduling strategies so that the target construction device operates with the corresponding control instruction. The method has the advantages that through the combination of prediction and model optimization, the working requirements can be predicted in advance, the scheduling strategy is optimized and adjusted, and the working efficiency and the resource utilization rate of the mine working equipment are improved.

In one possible implementation of the present application, as shown in fig. 8 to 10, the third sub-module 23 may include:

(1) As shown in fig. 8, the third sub-module 23 includes:

A convolutional neural network 231 for performing convolutional calculation on the historical operation data of the target construction equipment and the historical operation data of at least one first construction equipment, and outputting first data;

Namely, S21 in the intelligent scheduling method of the construction equipment shown in fig. 2 described above is performed. The specific implementation process is not described herein.

The first GRU network 232 is configured to process the first data input, and output first predicted time sequence data of the target construction device within a target preset duration;

Namely, S22 in the intelligent scheduling method of the construction equipment shown in fig. 2 described above is performed. The specific implementation process is not described herein.

An autoregressive model 234, configured to predict historical operation data of the target construction equipment, and output second predicted time-series data of the target construction equipment within a target preset duration;

Namely S23 in the intelligent scheduling method of the construction equipment shown in fig. 2 described above is performed. The specific implementation process is not described herein.

The first full connection layer 233 is configured to input the first predicted time-series data and the second predicted time-series data into the first full connection layer 233 for fusion, and output the fused time-series data.

Namely, S24 in the intelligent scheduling method of the construction equipment shown in fig. 2 described above is performed. The specific implementation process is not described herein.

(2) As shown in fig. 9, the third sub-module 23 includes:

Namely S210 in the intelligent scheduling method of the construction equipment shown in fig. 3 described above is performed. The specific implementation process is not described herein.

Namely, S220 in the intelligent scheduling method of the construction equipment shown in fig. 3 described above is performed. The specific implementation process is not described herein.

The second GRU network 235 is configured to process the first predicted time-series data and output third predicted time-series data.

Wherein the second GRU network is a periodically hopped GRU network. I.e., S230 in the intelligent scheduling method of the construction equipment shown in fig. 3 described above is performed. The specific implementation process is not described herein.

Namely, S240 in the intelligent scheduling method of the construction equipment shown in fig. 3 described above is performed. The specific implementation process is not described herein.

The first full connection layer 233 is configured to fuse the first predicted time series data, the second predicted time series data, and the third predicted time series data, and output the fused time series data.

I.e., S250 in the intelligent scheduling method of the construction equipment shown in fig. 3 described above is performed. The specific implementation process is not described herein.

(3) As shown in fig. 10, the third sub-module 23 includes:

namely, S25 in the intelligent scheduling method of the construction equipment shown in fig. 4 described above is performed. The specific implementation process is not described herein.

Namely, S26 in the intelligent scheduling method of the construction equipment shown in fig. 4 described above is performed. The specific implementation process is not described herein.

The second GRU network 235 processes the first predicted time-series data and outputs third predicted time-series data.

Wherein the second GRU network 235 is a cycle-hopped GRU network. Namely, S27 in the intelligent scheduling method of the construction equipment shown in fig. 4 described above is performed. The specific implementation process is not described herein.

The first full connection layer 233 is configured to fuse the first predicted time-series data and the third predicted time-series data, and output the fused time-series data.

Namely, S28 in the intelligent scheduling method of the construction equipment shown in fig. 4 described above is performed. The specific implementation process is not described herein.

In one possible implementation manner of the present application, the historical job data further includes historical space-time sequence data, where the historical space-time sequence data includes time points arranged according to a time sequence and corresponding space positions at the time points, and as shown in fig. 11, the intelligent scheduling system, the model base 2 further includes:

A spatial prediction model 25 for inputting the historical spatial time series data of the target construction equipment and the historical spatial time series data of at least one first construction equipment into the spatial prediction model to perform prediction, thereby generating predicted spatial time series data;

i.e. the spatial prediction model 25 is used to perform S6 shown in fig. 5 described above.

The fourth sub-module 24 is configured to fuse the historical job data, the fusion time series data, and the predicted space time series data of the target construction device, and generate predicted time series data of the target construction device within a target preset duration.

I.e. the fourth sub-module 24 is now used to perform S7 as shown in fig. 5 above.

In one possible implementation of the present application, as shown in fig. 12, the operation optimization model 3 further includes:

The real-time scheduling model 34, the real-time scheduling model 34 is used for detecting real-time sequence data of the target construction equipment within the target preset duration in real time; and when the real-time sequence data of the target construction equipment in the target preset time length is different from the predicted time sequence data of the target construction equipment, calling an operation optimization solver to adjust a scheduling strategy according to the real-time sequence data of the target construction equipment.

I.e. the real-time scheduling model 34 is used to perform S8 shown in fig. 6 described above. And when the real-time sequence data of the target construction equipment within the target preset time length is different from the predicted time sequence data of the target construction equipment, calling the operation optimization solver 33, so that the operation optimization solver 33 solves again according to the real-time sequence data of the target construction equipment to adjust the scheduling strategy.

Exemplary electronic device

Next, an electronic device according to an embodiment of the present application is described with reference to fig. 13.

Fig. 13 illustrates a block diagram of an electronic device according to an embodiment of the application.

As shown in fig. 13, the electronic device 60 includes one or more processors 600 and memory 602.

The processor 600 may be a Central Processing Unit (CPU) or other form of processing unit having data processing and/or instruction execution capabilities, and may control other components in the electronic device 60 to perform desired functions.

The memory 602 may include one or more computer program products that may include various forms of computer-readable storage media, such as volatile memory and/or non-volatile memory. The volatile memory may include, for example, random Access Memory (RAM) and/or cache memory (cache), and the like. The non-volatile memory may include, for example, read Only Memory (ROM), hard disk, flash memory, and the like. One or more computer program instructions may be stored on the computer readable storage medium that may be executed by the processor 11 to implement the simulation and optimization methods of the manufacturing model and/or other desired functions of the various embodiments of the present application described above.

In one example, the electronic device 60 may further include: an input device 601 and an output device 603, which are interconnected by a bus system and/or other form of connection mechanism (not shown).

When the electronic device is a stand-alone device, the input means 601 may be a communication network connector for receiving the acquired input signals from the first device and the second device.

In addition, the input device 601 may also include, for example, a keyboard, a mouse, and the like.

The output device 603 may output various information including the determined distance information, direction information, and the like to the outside. The output means 603 may include, for example, a display, speakers, a printer, and a communication network and remote output devices connected thereto, etc.

Of course, only some of the components of the electronic device 60 that are relevant to the present application are shown in fig. 13 for simplicity, components such as buses, input/output interfaces, etc. are omitted. In addition, the electronic device 60 may include any other suitable components depending on the particular application.

As a third aspect of the present application, there is provided a computer-readable storage medium storing a computer program for executing the steps of:

In addition to the methods and apparatus described above, embodiments of the application may also be a computer program product comprising computer program information which, when executed by a processor, causes the processor to perform the steps in the simulation and optimization methods of a manufacturing model according to various embodiments of the application described in this specification.

The computer program product may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C++ or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device, partly on a remote computing device, or entirely on the remote computing device or server.

Furthermore, embodiments of the present application may also be a computer-readable storage medium, on which computer program information is stored, which, when being executed by a processor, causes the processor to perform the steps in the simulation and optimization method of a manufacturing model according to various embodiments of the present application.

A computer readable storage medium may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may include, for example, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium would include the following: an electrical connection having one or more wires, a portable disk, a hard disk, random Access Memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

The basic principles of the present application have been described above in connection with specific embodiments, but it should be noted that the advantages, benefits, effects, etc. mentioned in the present application are merely examples and not intended to be limiting, and these advantages, benefits, effects, etc. are not to be construed as necessarily possessed by the various embodiments of the application. Furthermore, the specific details disclosed herein are for purposes of illustration and understanding only, and are not intended to be limiting, as the application is not necessarily limited to practice with the above described specific details.

The block diagrams of the devices, apparatuses, devices, systems referred to in the present application are only illustrative examples and are not intended to require or imply that the connections, arrangements, configurations must be made in the manner shown in the block diagrams. As will be appreciated by one of skill in the art, the devices, apparatuses, devices, systems may be connected, arranged, configured in any manner. Words such as "including," "comprising," "having," and the like are words of openness and mean "including but not limited to," and are used interchangeably therewith. The terms "or" and "as used herein refer to and are used interchangeably with the term" and/or "unless the context clearly indicates otherwise. The term "such as" as used herein refers to, and is used interchangeably with, the phrase "such as, but not limited to.

It is also noted that in the apparatus, devices and methods of the present application, the components or steps may be disassembled and/or assembled. Such decomposition and/or recombination should be considered as equivalent aspects of the present application.

Claims

1. An intelligent scheduling method for construction equipment is characterized by comprising the following steps:

Invoking historical operation data of target construction equipment and historical operation data of at least one first construction equipment of the same equipment as the target construction equipment, wherein the historical operation data comprises a plurality of historical preset time lengths and historical time sequence data of the construction equipment within the historical preset time lengths, and the historical time sequence data comprises time points arranged according to time sequences and working condition data corresponding to the time points;

fusing the historical operation data of the target construction equipment with the historical operation data of at least one first construction equipment to generate fused time sequence data;

Inputting the historical operation data and the fusion time series data of the target construction equipment to a second full-connection layer for fusion, and generating predicted time series data of the target construction equipment within a target preset duration;

acquiring constraint conditions corresponding to a plurality of target construction devices in a target preset time length of scheduling, and calling an operation optimization solver to perform simulation calculation on predicted time sequence data of the plurality of target construction devices in the target preset time length based on the constraint conditions so as to output scheduling strategies of the plurality of target construction devices in the target preset time length; and

Generating control instructions of each target construction equipment based on the scheduling strategy, so that each target construction equipment executes a corresponding task with the corresponding control instructions;

The method for generating the fusion time series data comprises the steps of: inputting the historical operation data of the target construction equipment and the historical operation data of at least one first construction equipment into a convolutional neural network for convolutional calculation, and outputting first data; inputting the first data into a first GRU network for processing, and outputting first predicted time sequence data of the target construction equipment within a target preset duration; inputting the historical operation data of the target construction equipment into an autoregressive model for prediction, and outputting second prediction time sequence data of the target construction equipment within a target preset duration; inputting the first predicted time sequence data and the second predicted time sequence data into a first full-connection layer for fusion, and outputting fusion time sequence data; or (b)

Fusing the historical job data of the target construction equipment with the historical job data of at least one first construction equipment to generate fused time series data, wherein the fused time series data comprises: inputting the historical operation data of the target construction equipment and the historical operation data of at least one first construction equipment into a convolutional neural network for convolutional calculation, and outputting first data; inputting the first data into a first GRU network for processing, and outputting first predicted time sequence data of the target construction equipment within a target preset duration; inputting the first predicted time sequence data into a second GRU network for processing, and outputting third predicted time sequence data, wherein the second GRU network is a cycle skip GRU network; inputting the historical operation data of the target construction equipment into an autoregressive model for prediction, and outputting second prediction time sequence data of the target construction equipment within a target preset duration; inputting the first predicted time sequence data, the second predicted time sequence data and the third predicted time sequence data into a first full-connection layer for fusion, and outputting fusion time sequence data; or (b)

Fusing the historical job data of the target construction equipment with the historical job data of at least one first construction equipment to generate fused time series data, wherein the fused time series data comprises: inputting the historical operation data of the target construction equipment and the historical operation data of at least one first construction equipment into a convolutional neural network for convolutional calculation, and outputting first data; inputting the first data into a first GRU network for processing, and outputting first predicted time sequence data of the target construction equipment within a target preset duration; inputting the first predicted time sequence data into a second GRU network for processing, and outputting third predicted time sequence data, wherein the second GRU network is a cycle skip GRU network; and inputting the first predicted time sequence data and the third predicted time sequence data into a first full-connection layer for fusion, and outputting fusion time sequence data.

2. The intelligent scheduling method of claim 1, wherein the historical job data further comprises historical space-time series data, the historical space-time series data comprising time points arranged in a time series and corresponding spatial locations at the time points, wherein the intelligent scheduling method further comprises:

Inputting the historical space time series data of the target construction equipment and the historical space time series data of at least one first construction equipment into a space prediction model for prediction to generate predicted space time series data;

and fusing the historical operation data, the fusion time series data and the prediction space time series data of the target construction equipment to generate prediction time series data of the target construction equipment within a target preset duration.

3. The intelligent scheduling method according to claim 1, wherein after generating the control instruction of each target construction equipment based on the scheduling policy so that each target construction equipment performs a corresponding task with the corresponding control instruction, the intelligent scheduling method further comprises:

Real-time sequence data of target construction equipment in a target preset duration are detected in real time;

and when the real-time sequence data of the target construction equipment in the target preset time length is different from the predicted time sequence data of the target construction equipment, calling the operation optimization solver to adjust a scheduling strategy according to the real-time sequence data of the target construction equipment.

4. An intelligent scheduling system for construction equipment, comprising:

The database is used for storing the operation data of the construction equipment;

The model library comprises a first submodule, wherein the first submodule is used for calling historical operation data of target construction equipment; the second sub-module is used for calling historical operation data of at least one first construction device of the same type of equipment as the target construction device; the third submodule is used for fusing the historical operation data of the target construction equipment with the historical operation data of at least one first construction equipment to generate fused time sequence data; the fourth submodule comprises a second full-connection layer, and is used for fusing the historical operation data of the target construction equipment with the fused time sequence data to generate predicted time sequence data of the target construction equipment within a target preset duration; and

An operational optimization model, the operational optimization model comprising: the system comprises a constraint condition module, a day-ahead scheduling module and an operation planning optimization solver, wherein the constraint condition module is used for selecting constraint conditions; the day-ahead scheduling module is used for acquiring constraint conditions corresponding to a plurality of target construction devices in a scheduled target preset time period, calling the operation optimization solver to perform simulation calculation on predicted time sequence data of the plurality of target construction devices in the target preset time period based on the constraint conditions so as to output a scheduling strategy of the plurality of target construction devices in the target preset time period, and generating a control instruction of each target construction device based on the scheduling strategy so that each target construction device executes a corresponding task with the corresponding control instruction;

Wherein the third sub-module comprises:

The convolution neural network is used for carrying out convolution calculation on the historical operation data of the target construction equipment and the historical operation data of at least one first construction equipment and outputting first data;

the first GRU network is used for processing the first data and outputting first predicted time sequence data of the target construction equipment within a target preset duration; and/or

The second GRU network is used for processing the first predicted time sequence data and outputting third predicted time sequence data, wherein the second GRU network is a cycle skip GRU network; and/or

The autoregressive model is used for predicting the historical operation data of the target construction equipment and outputting second predicted time sequence data of the target construction equipment within a target preset duration;

And the first full-connection layer is used for fusing the first predicted time sequence data, the second predicted time sequence data and/or the third predicted time sequence data and outputting fused time sequence data.

5. The intelligent scheduling system of claim 4, wherein the operational optimization model further comprises:

The real-time scheduling model is used for detecting real-time sequence data of the target construction equipment in a target preset duration in real time; and when the real-time sequence data of the target construction equipment in the target preset time length is different from the predicted time sequence data of the target construction equipment, calling the operation optimization solver to adjust a scheduling strategy according to the real-time sequence data of the target construction equipment.

6. The intelligent scheduling system of claim 4, wherein the historical job data further comprises historical space-time series data, the historical space-time series data comprising time points arranged in a time series and corresponding spatial locations at the time points;

wherein the model library further comprises:

The spatial prediction model is used for predicting the historical spatial time series data of the target construction equipment and the historical spatial time series data of at least one first construction equipment to generate predicted spatial time series data;

The fourth submodule is used for fusing the historical operation data of the target construction equipment, the fusion time series data and the prediction space time series data to generate prediction time series data of the target construction equipment within a target preset duration.