A shipyard micro-grid load interval prediction method based on process disturbance propagation and mode time shift correction

Abstract

The application discloses a shipyard micro-grid load interval prediction method based on process disturbance propagation and mode time shift correction, comprising: establishing an equipment characteristic library, disassembling various energy-consuming equipment operation of a shipyard into standby, start, stable operation, shutdown and other basic work modes, determining mode power curves through a parameterized power model, and forming the characteristic library; analyzing a production plan, outputting a process and equipment task sequence, extracting process constraints, equipment sharing, regional load conflict and auxiliary load linkage correlation; constructing a process disturbance event library, structurally defining four types of disturbances, namely process delay, advance, interruption and concurrency, setting trigger conditions and time of various disturbances, power correction parameters; deducing disturbance propagation combined with various correlation constraints to obtain equipment task disturbance correction; calling the characteristic library to generate multi-disturbance scenario equipment power curves; and outputting time-sharing load prediction intervals combined with reference load and offset load under different disturbances. The application can reduce the interference of disturbances on load prediction.

Description

Technical Field

[0001] This invention relates to the field of power grid load forecasting technology, and in particular to a method for forecasting load intervals in shipyard microgrids based on process disturbance propagation and mode time shift correction. Background Technology

[0002] Microgrid optimal dispatch is highly dependent on the accuracy of load-side input. Existing microgrid load forecasting methods typically fall into two categories: one uses a fixed average power value, typical daily curves, or a preset time period power sequence as the total system load input; the other extrapolates and forecasts future loads based on historical total load data, combined with statistical analysis, time series analysis, machine learning, or deep learning methods. The former is simple to model, but it usually simplifies the load to a given discrete power value sequence, making it difficult to reflect the dynamic evolution of the load; the latter, while able to characterize macroscopic fluctuation trends, is essentially still a data-driven historical extrapolation, lacking an explicit expression of equipment operating mechanisms and load formation processes.

[0003] For shipyard microgrids, the total load is formed by the superposition of the operating processes of major energy-consuming objects such as gantry cranes, dock pumps, large welding machines, air compressors, painting equipment, ventilation equipment, and auxiliary loads under different process tasks, exhibiting significant production task-driven and highly impactful characteristics. Existing methods, even those generating load curves based on deterministic production plans, typically assume that processes are executed in a predetermined sequence, making it difficult to account for disturbances such as process delays, advances, interruptions, and changes in concurrency relationships during actual production. These disturbances propagate along process dependencies, equipment sharing, regional conflicts, and auxiliary load linkages, further causing time shifts and changes in the overlap of equipment operating modes, leading to time-series shifts and local shape changes in the total load curve. Existing methods typically struggle to effectively characterize this process, thus failing to output interval-based load results reflecting the impact of disturbances and failing to meet the actual needs of optimized scheduling in shipyard microgrids. Summary of the Invention

[0004] To address the shortcomings of the existing technology, this invention provides a method for predicting the load range of a shipyard microgrid based on process disturbance propagation and mode time shift correction, which solves the problem that the predicted load curve is prone to deviation due to disturbances in the actual production process, making it difficult to meet the needs of microgrid optimal scheduling.

[0005] The technical solution of the present invention is as follows:

[0006] A method for predicting load intervals in shipyard microgrids based on process disturbance propagation and mode time shift correction includes:

[0007] Step S1: Establish equipment feature library: Decompose the operation of each type of energy-consuming equipment in the shipyard into multiple basic operating modes. Each basic operating mode uniquely determines its power curve characteristics through a parameterized power model. The basic operating modes include at least standby mode, start-up mode, stable operating mode and shutdown mode. The basic operating modes of all equipment and their corresponding parameterized power models constitute the equipment feature library.

[0008] Step S2, Production Plan Parsing: Obtain the production plan for future time periods, parse the production plan into process task sequences and equipment task sequences, and extract the constraints before and after the processes, equipment sharing relationships, regional conflict relationships, and auxiliary load linkage relationships;

[0009] Step S3: Construct a process disturbance event library: Structurally represent the four types of disturbance events that may occur during the actual production process of the shipyard: process delay, process advance, process interruption, and process concurrency. Define triggering conditions and corresponding time-type and power-type correction parameters for each type of disturbance.

[0010] Step S4, Process Disturbance Propagation Analysis: Based on the process-related constraints, equipment sharing relationships, regional conflict relationships, and auxiliary load linkage relationships, the process disturbance events are propagated and deduced to obtain the disturbance correction results for equipment tasks;

[0011] Step S5, Mode Time Shift Correction and Power Curve Generation: Based on the disturbance correction results, the start and end time shifts, duration corrections, and overlap adjustments are made to the sequence of primitive working modes corresponding to the device, and the device feature library is called to generate the predicted power curves of each device under different disturbance scenarios.

[0012] Step S6: Generate load intervals: Accumulate the corrected power curves of all devices under the same disturbance scenario on a unified time axis to form the total load curve under that scenario; generate time-sharing load intervals for future periods using the baseline load curve and the total load curves under different disturbance scenarios.

[0013] Furthermore, in step S1, the parameters of the parameterized power model include at least one of standby power, rated power, duration, power rise rate, and power fall rate.

[0014] Furthermore, the pre- and post-process constraint relationship is used to characterize the sequential execution dependency relationship between processes; the equipment sharing relationship is used to characterize the occupation relationship of multiple process tasks on the same equipment resource; the regional conflict relationship is used to characterize the constraint relationship that multiple process tasks cannot be executed simultaneously in the same region; and the auxiliary load linkage relationship is used to characterize the correlation relationship that the auxiliary load changes synchronously with the start, operation or stop of the main process task.

[0015] Furthermore, the process disturbance events include process delays, process advancements, process interruptions, changes in process duration, changes in process concurrency relationships, and timing disturbances caused by temporarily inserted tasks.

[0016] Furthermore, the time-related correction parameters include process start time offset, end time offset, duration correction, and concurrency relationship correction.

[0017] Furthermore, the power correction parameters include power value correction, power ramp-up rate correction, and power decay rate correction.

[0018] Furthermore, the propagation deduction in step S4 specifically includes: propagating the disturbance of the preceding process to the subsequent process based on the process constraint relationship; propagating the current equipment occupancy disturbance to the subsequent shared equipment task based on the equipment sharing relationship; propagating the area occupancy disturbance to the related process task in the same area based on the area conflict relationship; and propagating the main process task disturbance to the associated auxiliary load task based on the auxiliary load linkage relationship.

[0019] Furthermore, in step S5, the start and end times of the sequence of basic operating modes corresponding to the device are offset, the duration is corrected, and the overlap relationship is adjusted. Specifically, this includes: shifting and correcting the start time of the basic operating mode; shifting and correcting the end time of the basic operating mode; scaling and correcting the duration of the basic operating mode; and rearranging and correcting the overlap relationship between multiple basic operating modes. The rearrangement correction involves adjusting the time overlap interval between multiple basic operating modes to change the load superposition relationship of each basic operating mode.

[0020] Furthermore, in step S6, the total load curves under different disturbance scenarios include an upper-biased load curve where the total load is higher than the load level corresponding to the reference load curve and a lower-biased load curve where the total load is lower than the load level corresponding to the reference load curve. The time-sharing load interval is the area between the upper-biased load curve and the lower-biased load curve.

[0021] Furthermore, the upper-biased load curve is the boundary curve formed by the larger value of the total load at each time under different disturbance scenarios, and the lower-biased load curve is the boundary curve formed by the smaller value of the total load at each time under different disturbance scenarios.

[0022] Compared with the prior art, the present invention has the following advantages:

[0023] This invention constructs a process disturbance event library and combines process-related constraints, equipment sharing relationships, regional conflict relationships, and auxiliary load linkage relationships to propagate and deduce process disturbances. Based on the baseline load curve, it can further generate upper-biased load curves, lower-biased load curves, and time-sharing load intervals, thereby expanding the load prediction results from a single definite value to interval-based results, which better reflects the actual fluctuation characteristics of the shipyard production process and improves the reliability and applicability of the prediction results.

[0024] This invention constructs an equipment feature library based on equipment element operating modes and parameterized power models, and maps the propagation results of process disturbances to the start and end time offsets, duration corrections, and overlap adjustments of equipment mode sequences. This allows it to reflect the changes in load curves on the time axis, such as forward shifts, backward shifts, broadening, compression, and local peak reconstruction. This method not only improves the precision of load modeling for major energy-consuming equipment in shipyards but also enhances the physical interpretability of the prediction results.

[0025] The baseline load curve, upper-biased load curve, lower-biased load curve, and time-of-use load interval generated by this invention can provide shipyard microgrids with richer and more effective load-side information in areas such as day-ahead scheduling, intraday rolling scheduling, energy storage charging and discharging strategy formulation, demand peak early warning, and risk-constrained operation. Compared to methods that only provide a single forecast curve, this invention enables the scheduling system to identify potential high-load risk windows and low-load idle periods in advance, providing a basis for energy storage capacity reservation, load peak shaving and valley filling, and flexible resource coordination control, thereby improving the foresight and practicality of microgrid operation optimization in shipyards and other high-energy-consuming manufacturing scenarios. Attached Figure Description

[0026] Figure 1 The flowchart illustrates a method for predicting the load range of a shipyard microgrid based on process disturbance propagation and mode time shift correction, as an example.

[0027] Figure 2 This is a schematic diagram of the device's basic operating modes and parameterized power model.

[0028] Figure 3 This is a schematic diagram illustrating the propagation relationship of process disturbances.

[0029] Figure 4 This is a schematic diagram of equipment mode time shift correction before and after process disturbance. Detailed Implementation

[0030] The present invention will be further described below with reference to embodiments, but these are not intended to limit the scope of the invention.

[0031] Please combine Figure 1As shown, this embodiment takes a typical segmented construction scenario in a shipyard as an example to illustrate a method for predicting the load range of a shipyard microgrid based on process disturbance propagation and mode time shift correction. The method includes the following steps:

[0032] Step 1: Establish an Equipment Feature Library. The operation of each type of energy-consuming equipment in the shipyard is decomposed into multiple basic operating modes. Each basic operating mode uniquely determines its power curve characteristics through a parametric power model. The basic operating modes of the equipment and their corresponding parametric power models constitute the equipment feature library. The main energy-consuming equipment in the shipyard includes hoisting equipment, welding equipment, air compressor equipment, painting equipment, ventilation equipment, transfer equipment, and lighting equipment. Each type of equipment has different operating characteristics, but the overall basic operating modes can include standby mode, start-up mode, stable operating mode, and shutdown mode. Each basic operating mode can be uniquely determined by one or more key operating parameters. These key operating parameters can be standby power, rated power, duration, power rise rate, power fall rate, and task load factor, etc. The parametric power model is a function expressing the change of equipment power over time based on these parameters.

[0033] like Figure 2 The diagram shows the basic operating modes and parameterized power model of the welding machine. In standby mode M0, the welding machine is powered on and in standby mode, with the power maintained at a low level. The standby power is P. idle (kW); In start-up mode M1, welding starts from standby state, and the power gradually increases until it stabilizes. This stage can be controlled by the power increase rate r. up (kW / min) Description: In stable operating mode, i.e., steady-state welding mode M2, the welding machine welds stably at its rated power, maintaining a high power level. This stage is driven by the rated power P. r (kW) and steady-state welding duration T d (min) Description: In shutdown mode M3, welding is completed and the power gradually decreases to the standby level. This stage is measured by the power increase rate r. down (kW / min) Description. From this, the parameterized power model of the welding machine is obtained: P(t;θ)=f(t;P r ,P idle ,T d ,r up ,r down ), where θ is the parameter combination θ={P r ,P idle ,T d ,r up ,r downIt can be seen that for welding, the power of the equipment under each basic operating mode changes as a linear function over time. It should be noted that this example only demonstrates one feasible parametric power model for a welding machine. For other types of equipment, different power variation patterns may exist under different basic operating modes, resulting in parametric power models different from those for the welding machine.

[0034] Step 2, Production Plan Analysis: At each rolling time point, obtain a refined production plan for future time periods. This plan includes the specific tasks and working hours for each piece of equipment. The production plan needs to be analyzed into a sequence of process tasks and a sequence of equipment tasks, and the pre- and post-process constraints, equipment sharing relationships, regional conflict relationships, and auxiliary load linkage relationships need to be extracted. Pre- and post-process constraints characterize the sequential execution dependencies between processes; equipment sharing relationships characterize the resource occupancy of multiple process tasks on the same equipment; regional conflict relationships characterize the constraints that prevent multiple process tasks from executing simultaneously within the same region; and auxiliary load linkage relationships characterize the synchronous changes in auxiliary loads as the main process tasks start, run, or stop.

[0035] A process task sequence primarily describes the sequence of tasks performed. For example, in a shipyard, a certain section might require multiple tasks such as "welding," "hoisting," and then "welding" again. The process task sequence is a production process flow that reflects the logical relationships between processes.

[0036] Based on this, each process task in the production plan is analyzed into an equipment task sequence according to time sequence. For example, welding tasks are analyzed into the start-stop mode of the welding machine, and hoisting tasks are analyzed into the operation mode of the gantry crane. According to historical data and equipment process requirements, the key parameter values ​​of the parametric power model of the equipment used in each task are inferred, such as the rated power and duration in the parametric power model of welding established above. With these parameters, the power demand of the equipment in future periods can be predicted more accurately.

[0037] Step 3: Construct a process disturbance event library: Structure and represent the process disturbance events that may occur during actual shipyard production. Define trigger conditions and corresponding time-based and power-based correction parameters for each type of disturbance. Process disturbance events include process delays, process advancements, process interruptions, changes in process duration, changes in process concurrency relationships, and timing disturbances caused by temporarily inserted tasks. Time-based correction parameters include process start time offset, end time offset, duration correction, and concurrency relationship correction. Specifically, the process start time offset corrects the deviation of the target process's actual start time from the planned schedule; the end time offset corrects the deviation of the target process's actual completion time from the planned end time; the duration correction corrects the deviation between the actual execution time of the target process and the planned process time; and the concurrency relationship correction corrects the parallel execution relationship, overlap, or timing coupling relationship between multiple processes.

[0038] As shown in Table 1, common process disturbances include process delays, process advances, process interruptions, and process concurrency. Each disturbance affects the load curve. First, we establish the disturbance triggering conditions: each disturbance event has its own triggering conditions. For example, process delays may be triggered by equipment failures or material shortages; process advances may be triggered by production line optimization or efficiency improvements. Next, we determine the disturbance correction parameters: based on the actual situation, we determine the correction parameters for each disturbance. For the parameterized power model, the main influencing parameters are power and time. Therefore, the correction parameters mainly include time-related correction parameters and power-related correction parameters. For example, material shortages causing process delays affect the welding machine's startup mode, resulting in a shift in the start time. The time-related correction parameter is to shift the start time backward by Δt. Similarly, process advances caused by increased production efficiency affect the welding machine's startup mode, also resulting in a shift in the start time. The time-related correction parameter is to advance the startup time, while the power-related correction parameter is to increase the power ramp-up rate.

[0039] Table 1. Examples of Correction Rules for Process Disturbance Parameters

[0040]

[0041] Step 4: Process disturbance propagation analysis: Based on the constraints before and after the process, the equipment sharing relationship, the regional conflict relationship, and the auxiliary load linkage relationship, the process disturbance event is propagated and deduced to obtain the disturbance correction result of the equipment task.

[0042] The specific content of the communication simulation includes:

[0043] There may be time or equipment resource constraints between certain tasks. For example, a welding task can only begin after a hoisting task is completed, or certain equipment can only be used in a specific area. It is necessary to infer the scope of the disturbance by propagating the disturbance from the preceding process to the subsequent process based on the constraints between processes.

[0044] In some cases, multiple devices may share the same time period or resources. For example, multiple welding machines may use the same power supply simultaneously, or multiple hoisting devices may need to use the same hoisting equipment simultaneously. The sharing relationship between devices can also affect the propagation of disturbances, requiring the current device occupancy disturbance to be propagated to subsequent shared device tasks based on the device sharing relationship.

[0045] In shipyards, different processes may conflict spatially; for example, multiple processes may be performed in the same location, preventing equipment from operating simultaneously. The impact of these regional conflicts on load demand needs to be considered. Based on the regional conflict relationships, regional occupancy disturbances are propagated to related process tasks within the same region.

[0046] Some auxiliary equipment, such as fans and air conditioners, may vary depending on the operating status of the main equipment. For example, when welding equipment is in use, the load on the fans may increase. It is necessary to consider the impact of these interrelationships on the load and, based on the auxiliary load interrelationships, propagate disturbances from the main process task to related auxiliary load tasks.

[0047] like Figure 3 As shown, the results of the disturbance propagation analysis for one process are as follows: Process A is segmented hoisting, which experiences a delay disturbance. The original planned working time was 08:00-08:30, but after the disturbance, it is adjusted to 08:20-08:50. Process A2 is hoisting task 2. Since it shares a gantry crane with Process A, it belongs to the equipment sharing relationship. The original planned working time of Process A2 was 08:30-09:00, which needs to be extended accordingly. After the disturbance, it is adjusted to 08:50-09:20. Process B: Positioning welding is a subsequent process of Process A, and Process C: Repair welding is a subsequent process of Process B. Therefore, there is a constraint relationship between the preceding and following processes. The original planned working time of Process B was 08:30-10:00, and the original planned working time of Process C was 10:00-12:00. The corresponding adjustments are 08:50-10:20 and 10:20-12:20. Process D is a painting operation, which shares the same area as Process A2, creating an area conflict. The original planned working time for Process D was 09:00-11:00, which is adjusted to 09:20-11:20. Auxiliary load E is ventilation equipment, which needs to be started / stopped synchronously with Process D. Processes D and E have an auxiliary load linkage relationship; therefore, the original planned working time for auxiliary load E, 09:00-11:00, is also adjusted to 09:20-11:20.

[0048] Step 5: Based on the disturbance correction results, perform start and end time offset, duration correction and overlap adjustment on the corresponding primitive working mode sequence of the device, and call the device feature library to generate the predicted power curve of each device under different disturbance scenarios.

[0049] Specific modifications to the primitive operating mode sequence include: shifting and correcting the start time of the primitive operating mode; shifting and correcting the end time of the primitive operating mode; scaling and correcting the duration of the primitive operating mode; and rearranging and correcting the overlapping relationships between multiple primitive operating modes. Rearranging and correcting involves adjusting the time overlap intervals between multiple primitive operating modes to change the load superposition relationship of each primitive operating mode.

[0050] For example, process delays may postpone equipment startup, process advancement may shorten equipment uptime, process interruptions may cause power drops leading to load breaks, and process concurrency may cause load overlap and instantaneous power increases. Generating corrected power curves: Based on the mode time shift correction, new power curves for the equipment are generated after the disturbance. These corrected power curves will reflect the possible changes in load after the disturbance.

[0051] like Figure 4 As shown, when a delay disturbance occurs in the process, the welding process is delayed by Δt due to the influence of the preceding process, and the welding operation time is extended by Δd. Therefore, the start time of the welding machine's standby mode, start mode, and steady-state welding mode is delayed by Δt, while the start time of the shutdown mode is delayed by Δt+Δd.

[0052] Step 6: Generate Load Range: Accumulate the corrected power curves of each device on the time axis to obtain the total load curve of the shipyard microgrid for future periods. This total load curve reflects the load demand of all devices under disturbance conditions. Generate Load Range: By comparing the baseline load curve, the upper-biased load curve, and the lower-biased load curve, generate the load range (the shaded area between the upper-biased load and the lower-biased load). At a certain moment, the welding machine has not yet started operation in the baseline plan, while the gantry is in normal operation. If, after considering process disturbances, the welding machine is put into operation ahead of schedule due to the compression of the preceding process, and the gantry is delayed and temporarily not in operation due to insufficient hoisting preparation, the welding machine is a high-power device, and the additional load brought by its early start is greater than the load reduction caused by the delay of the gantry. Therefore, the total load at this moment will be higher than the baseline load, forming an upper-biased load scenario. Conversely, in another disturbance scenario, the welding machine does not start ahead of schedule and is still started in a later period as originally planned, while the gantry is temporarily not in operation at the current moment due to the delay of the preceding hoisting task. At this point, the gantry load that should have been present at that moment is removed, while the welding machine load has not yet been generated. The total load is lower than the load level corresponding to the baseline load curve, forming a down-biased load scenario. The baseline load curve is the total load forecast curve generated according to the deterministic production plan and the original equipment task sequence; the up-biased load curve is the boundary curve formed by the larger value of the total load at each moment under different disturbance scenarios; the down-biased load curve is the boundary curve formed by the smaller value of the total load at each moment under different disturbance scenarios. Determining the load range at a certain moment can clarify the range of load changes under the condition of disturbance, providing support for subsequent microgrid scheduling and energy storage optimization.

Claims

1. A method for predicting load intervals in shipyard microgrids based on process disturbance propagation and mode time shift correction, characterized in that, include: Step S1: Establish equipment feature library: Decompose the operation of each type of energy-consuming equipment in the shipyard into multiple basic operating modes. Each basic operating mode uniquely determines its power curve characteristics through a parameterized power model. The basic operating modes include at least standby mode, start-up mode, stable operating mode and shutdown mode. The basic operating modes of all equipment and their corresponding parameterized power models constitute the equipment feature library. Step S2, Production Plan Parsing: Obtain the production plan for future time periods, parse the production plan into process task sequences and equipment task sequences, and extract the constraints before and after the processes, equipment sharing relationships, regional conflict relationships, and auxiliary load linkage relationships; Step S3: Construct a process disturbance event library: Structurally represent the four types of disturbance events that may occur during the actual production process of the shipyard: process delay, process advance, process interruption, and process concurrency. Define triggering conditions and corresponding time-type and power-type correction parameters for each type of disturbance. Step S4, Process Disturbance Propagation Analysis: Based on the process-related constraints, equipment sharing relationships, regional conflict relationships, and auxiliary load linkage relationships, the process disturbance events are propagated and deduced to obtain the disturbance correction results for equipment tasks; Step S5, Mode Time Shift Correction and Power Curve Generation: Based on the disturbance correction results, the start and end time shifts, duration corrections, and overlap adjustments are made to the sequence of primitive working modes corresponding to the device, and the device feature library is called to generate the predicted power curves of each device under different disturbance scenarios. Step S6: Generate load intervals: Accumulate the corrected power curves of all devices under the same disturbance scenario on a unified time axis to form the total load curve under that scenario; generate time-sharing load intervals for future periods using the baseline load curve and the total load curves under different disturbance scenarios.

2. The method for predicting the load range of a shipyard microgrid based on process disturbance propagation and mode time shift correction as described in claim 1, characterized in that, In step S1, the parameters of the parameterized power model include at least one of standby power, rated power, duration, power rise rate, and power fall rate.

3. The method for predicting the load range of a shipyard microgrid based on process disturbance propagation and mode time shift correction as described in claim 1, characterized in that, The pre- and post-process constraints are used to characterize the sequential execution dependencies between processes. The equipment sharing relationship is used to characterize the occupation relationship of multiple process tasks on the same equipment resource; the regional conflict relationship is used to characterize the constraint relationship that multiple process tasks cannot be executed simultaneously in the same region; the auxiliary load linkage relationship is used to characterize the association relationship that the auxiliary load changes synchronously with the start, operation or stop of the main process task.

4. The method for predicting the load range of a shipyard microgrid based on process disturbance propagation and mode time shift correction according to claim 1, characterized in that, The process disturbance events include process delays, process advancements, process interruptions, changes in process duration, changes in process concurrency relationships, and timing disturbances caused by temporarily inserted tasks.

5. The method for predicting the load range of a shipyard microgrid based on process disturbance propagation and mode time shift correction according to claim 4, characterized in that, The time-related correction parameters include process start time offset, end time offset, duration correction, and concurrency relationship correction.

6. The method for predicting the load range of a shipyard microgrid based on process disturbance propagation and mode time shift correction according to claim 4, characterized in that, The power correction parameters include power value correction, power ramp-up rate correction, and power decay rate correction.

7. The method for predicting the load range of a shipyard microgrid based on process disturbance propagation and mode time shift correction according to claim 1, characterized in that, The propagation and deduction in step S4 specifically includes: propagating the disturbance of the preceding process to the subsequent process based on the process constraint relationship; propagating the current equipment occupancy disturbance to the subsequent shared equipment task based on the equipment sharing relationship; propagating the area occupancy disturbance to the related process task in the same area based on the area conflict relationship; and propagating the main process task disturbance to the associated auxiliary load task based on the auxiliary load linkage relationship.

8. The method for predicting the load range of a shipyard microgrid based on process disturbance propagation and mode time shift correction according to claim 1, characterized in that, In step S5, the start and end times of the corresponding basic element working mode sequence of the device are offset, the duration is corrected, and the overlap relationship is adjusted. Specifically, this includes: shifting and correcting the start time of the basic element working mode; shifting and correcting the end time of the basic element working mode; scaling and correcting the duration of the basic element working mode; and rearranging and correcting the overlap relationship between multiple basic element working modes. The rearrangement correction is to adjust the time overlap interval between multiple basic element working modes to change the load superposition relationship of each basic element working mode.

9. The method for predicting the load range of a shipyard microgrid based on process disturbance propagation and mode time shift correction according to claim 1, characterized in that, In step S6, the total load curves under different disturbance scenarios include an upper-biased load curve where the total load is higher than the load level corresponding to the reference load curve, and a lower-biased load curve where the total load is lower than the load level corresponding to the reference load curve. The time-sharing load interval is the area between the upper-biased load curve and the lower-biased load curve.

10. The method for predicting the load range of a shipyard microgrid based on process disturbance propagation and mode time shift correction according to claim 9, characterized in that, The upper-biased load curve is the boundary curve formed by the larger value of the total load at each time under different disturbance scenarios, and the lower-biased load curve is the boundary curve formed by the smaller value of the total load at each time under different disturbance scenarios.

Images