Port multi-energy collaborative control system and method based on peak-valley electricity price and load prediction

Abstract

This invention discloses a port multi-energy coordinated control system and method based on peak-valley electricity pricing and load forecasting, relating to the field of port energy conservation and new energy utilization technology. It includes constructing a multi-energy coupled system composed of photovoltaic power generation modules, wind power generation modules, gravity energy storage modules, supercapacitor energy storage modules, DC bus modules, and energy management modules, and realizing multi-source energy collection and distribution through the DC bus; based on meteorological data, operational data, and historical data, using an LSTM-Attention model for multi-timescale load and output forecasting; combining time-of-use electricity pricing and energy storage status, using mixed-integer linear programming to generate an optimal scheduling plan, and using model predictive control to continuously correct the scheduling results. This invention establishes a coordinated control mechanism to achieve dynamic optimization scheduling of the multi-energy system, improve the renewable energy absorption capacity, reduce operating costs, and enhance the stability and flexibility of system operation.

Description

Technical Field

[0001] This invention relates to the field of port energy conservation and new energy utilization technology, and in particular to a port multi-energy coordinated control system and method based on peak-valley electricity prices and load forecasting. Background Technology

[0002] Port energy systems are gradually evolving from traditional single-grid power supply to a multi-energy integration and synergistic optimization model. Currently, renewable energy sources such as photovoltaic (PV) and wind power are increasingly widely used in port scenarios, while energy storage technologies (such as electrochemical energy storage and supercapacitor energy storage) are also increasingly being used for peak shaving and valley filling, and energy regulation. However, due to the significant randomness and time-varying nature of port operations, their load is affected by multiple factors such as ship arrival time, operational intensity, and equipment operating status, exhibiting strong fluctuations and uncertainties. Furthermore, PV and wind power output is intermittent and random, making it difficult for traditional energy management methods based on empirical rules or static scheduling to achieve efficient matching and synergistic utilization of multiple energy sources. In existing technologies, most systems only conduct scheduling for a single energy storage form or a single optimization objective, lacking global optimization capabilities oriented towards peak-valley pricing mechanisms. They also have shortcomings in multi-timescale forecasting and rolling correction, leading to low renewable energy absorption rates, high system operating costs, and poor energy utilization efficiency. Therefore, how to achieve dynamic synergistic control of port multi-energy systems based on time-of-use pricing and load forecasting has become a key technical problem that urgently needs to be solved.

[0003] CN116826862A discloses a method for coordinated scheduling of multi-energy flow and container logistics in an electric-hydrogen port. This method establishes a coupling relationship between the electric-hydrogen energy system and the port logistics system, and builds a multi-objective optimization model with the goals of minimizing operating costs and optimizing logistics efficiency. It combines an improved non-dominated sorting genetic algorithm and the Benders decomposition method to achieve joint scheduling. While this method improves port energy utilization efficiency and logistics coordination to some extent, it focuses on the coordinated optimization of the electric-hydrogen energy system and the logistics system. It does not deeply model the output prediction and uncertainty correction for fluctuating power sources such as photovoltaic and wind power, nor does it consider the optimization of energy storage charging and discharging strategies driven by peak-valley electricity prices. Furthermore, it lacks multi-timescale prediction and dynamic rolling correction mechanisms, resulting in insufficient adaptability and real-time performance under complex actual operating conditions.

[0004] CN115983746B proposes a multi-agent collaborative optimization method for port operation scheduling. By constructing an optimization model with operation time and cost as objectives and introducing the Grey Wolf algorithm for solution, it achieves collaborative scheduling among multiple entities such as berths, quay cranes, and AGVs. This method primarily focuses on operational process optimization and resource allocation efficiency improvement. While it considers energy consumption and carbon emission factors to some extent, it does not address the physical coupling of multi-energy systems and energy flow scheduling issues. It lacks a systematic design for collaborative control of energy storage systems and renewable energy consumption mechanisms, and in particular, it fails to establish an energy management strategy based on load forecasting and electricity price signals, making it difficult to achieve collaborative optimization between the energy side and the load side. Summary of the Invention

[0005] The purpose of this section is to outline some aspects of the embodiments of the present invention and to briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section, as well as in the abstract and title of the present application, to avoid obscuring the purpose of this section, the abstract and title of the invention. Such simplifications or omissions shall not be used to limit the scope of the present invention.

[0006] In view of the problems that existing port multi-energy system scheduling technologies generally suffer from, such as insufficient multi-energy coordination, lack of peak-valley electricity price guidance mechanism, disconnect between prediction and control, and low energy storage utilization efficiency, this invention is proposed.

[0007] Therefore, the problem to be solved by this invention is how to achieve efficient coordinated scheduling and dynamic optimization control among port photovoltaic power generation, wind power generation and multiple types of energy storage under the conditions of considering peak and valley electricity prices and multi-time scale load forecasting.

[0008] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, embodiments of the present invention provide a port multi-energy coordinated control method based on peak-valley electricity prices and load forecasting, comprising, Photovoltaic power generation modules are used to convert solar energy into electrical energy, which is then connected to a common DC bus via a DC converter. Wind power generation module, used to convert wind energy into electrical energy, which is then connected to a common DC bus via an AC / DC converter; Gravity energy storage module is used to store potential energy by using standard port containers as the energy storage medium. The container is lifted to a predetermined height by a lifting mechanism and the potential energy is released to generate electricity by a lowering mechanism, realizing the bidirectional conversion of potential energy and electrical energy. The supercapacitor energy storage module is used to provide instantaneous power support, smooth system power fluctuations, recover regenerative braking energy generated during crane operation, and connect to the common DC bus through a bidirectional DC converter; The DC bus module is used as an energy collection and distribution center to realize DC grid connection of multi-source energy; The energy management module is used to monitor the operating status of each module in real time, adopts a predictive-correction two-layer architecture to formulate energy allocation strategies, and realizes coordinated control of each module; Among them, the photovoltaic power generation module, wind power generation module, gravity energy storage module, and supercapacitor energy storage module are all connected in parallel to the common DC bus of the DC bus module.

[0009] Secondly, embodiments of the present invention provide a port multi-energy coordinated control system based on peak-valley electricity prices and load forecasting, comprising: S1: Collect meteorological data, operational data, historical data and external data. After preprocessing the collected data, input it into the LSTM-Attention prediction model to perform long-term, medium-term and short-term load forecasts. The predicted output values ​​include photovoltaic power generation, wind power generation, port electricity consumption and grid power supply. S2: Based on long-term load forecasts, combined with time-of-use electricity price information and the initial state of charge of energy storage modules, with the goal of minimizing overall operating costs, mixed-integer linear programming is used to solve the forecast scheduling plan within the forecast time period, which includes energy storage charging and discharging plan, photovoltaic / wind power generation plan and grid interaction plan. S3: Based on the medium-term load forecast and forecast scheduling plan, combined with real-time electricity price, actual energy storage capacity and photovoltaic / wind power forecast errors, model predictive control is used to correct the scheduling plan within the forecast period.

[0010] Thirdly, embodiments of the present invention provide a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement any step of the above-described port multi-energy coordinated control method based on peak-valley electricity prices and load forecasting.

[0011] Fourthly, embodiments of the present invention provide a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the above-described port multi-energy coordinated control method based on peak-valley electricity prices and load forecasting.

[0012] Compared with existing technologies, the beneficial effects of this invention are as follows: Step S1 involves multi-source data collection from meteorological data, operational data, historical data, and external data. Combined with preprocessing and an LSTM-Attention model, load and output predictions are achieved at long, medium, and short time scales. This enables the system to obtain prediction results with temporal continuity and adaptive feature weights, providing high-confidence prior information support for subsequent scheduling. This process not only achieves quantitative modeling of port load fluctuations and the uncertainty of wind and solar power output but also effectively improves the foresight and interpretability of multi-energy system operation decisions, ultimately reducing the impact of prediction bias on system operation and improving overall scheduling accuracy. Step S2 introduces a time-of-use pricing mechanism and initial state constraints for energy storage based on long-term prediction results, constructing a mixed-integer linear programming model with the objective of minimizing comprehensive operating costs. This model enables predictions of photovoltaic, wind, supercapacitor, and gravity energy storage. The collaborative charging and discharging decision-making among energy sources and the grid interaction optimization configuration step unifies the modeling of energy supply and demand balance constraints and economic objectives. This not only achieves the global optimal allocation among multiple energy sources, but also strengthens the functional division of different energy storage at the time scale through a hierarchical scheduling strategy that prioritizes supercapacitors and then gravity energy storage. This significantly reduces electricity costs and improves the renewable energy absorption rate while ensuring system stability. Step S3 introduces a model predictive control mechanism based on medium-term forecast results, real-time electricity prices, and actual operating deviations. This mechanism performs rolling corrections to the established scheduling plan and combines deviation correction models for photovoltaic, wind power, and load to achieve dynamic closed-loop control. This enables the system to continuously correct prediction errors and adapt to external disturbances during actual operation. This process effectively compensates for the shortcomings of traditional static scheduling in terms of real-time performance and robustness, thereby improving system operational stability, enhancing resistance to uncertainty, and ensuring the continuous optimization of the scheduling strategy. Attached Figure Description

[0013] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein: Figure 1 This is a topology diagram of a port multi-energy coordinated control system based on peak-valley electricity prices and load forecasting. Figure 2 This is a flowchart of a port multi-energy coordinated control method based on peak-valley electricity prices and load forecasting. Detailed Implementation

[0014] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0015] Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without inventive effort should fall within the scope of protection of this invention.

[0016] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0017] As mentioned in the background section, existing technologies mostly focus on a single energy form or a single optimization objective, lacking a unified multi-source energy coupling architecture and a rolling correction mechanism based on prediction results, making it difficult to cope with the dual uncertainties of port load and renewable energy output. To address these issues, this invention provides a port multi-energy coordinated control method based on peak-valley electricity prices and load forecasting.

[0018] Reference Figure 1 , Figure 1 This is a flowchart illustrating a port multi-energy coordinated control method based on peak-valley electricity prices and load forecasting, according to an embodiment of the present invention. Figure 1 As shown, a port multi-energy collaborative control system based on peak-valley electricity prices and load forecasting includes: a photovoltaic power generation module, a wind power generation module, a gravity energy storage module, a supercapacitor energy storage module, a DC bus module, and an energy management module; wherein, the photovoltaic power generation module, the wind power generation module, the gravity energy storage module, and the supercapacitor energy storage module are all connected in parallel to the common DC bus of the DC bus module.

[0019] Specifically, the photovoltaic power generation module is used to convert solar energy into electrical energy, which is then connected to the common DC bus via a DC-DC converter. The DC-DC converter uses maximum power point tracking control, which enables the photovoltaic output power to be adaptively adjusted according to the DC bus voltage. The wind power generation module is used to convert wind energy into electrical energy, which is then connected to the common DC bus via an AC / DC converter. The wind power generation module is equipped with a maximum power point tracking controller, which dynamically adjusts the wind turbine speed and output power according to changes in wind speed.

[0020] In an optional embodiment, the DC converter uses a Boost converter circuit with an output voltage range of 1000-1500V; the AC / DC converter uses a three-phase PWM rectifier circuit; the total installed capacity of the photovoltaic power generation module is 500kWp, and the total installed capacity of the wind power generation module is 300kW.

[0021] Furthermore, the output power of the photovoltaic and wind power modules adaptively adjusts according to the DC bus voltage, including: when the DC bus voltage is higher than the rated voltage, the output power is reduced to suppress overvoltage; when the DC bus voltage is lower than the rated voltage, the maximum power is output to meet the load demand; when the sum of the output power of the photovoltaic and wind power modules is greater than the actual load demand, surplus power is calculated and distributed to each energy storage module for charging in a priority order of supercapacitor energy storage over gravity energy storage; when each energy storage module reaches its charging limit, the surplus power is fed back to the grid.

[0022] Furthermore, the gravity energy storage module uses standard port containers as the energy storage medium. The container is lifted to a predetermined height by a lifting mechanism to store potential energy, and the potential energy is released to generate electricity by a lowering mechanism, realizing the bidirectional conversion of potential energy and electrical energy. The supercapacitor energy storage module is used to provide instantaneous power support, smooth system power fluctuations, and recover the regenerative braking energy generated during crane operation. It is connected to the common DC bus through a bidirectional DC converter. The DC bus module is used as an energy collection and distribution center to realize DC grid connection of multi-source energy.

[0023] In an optional embodiment, the gravity energy storage module uses a standard port container with a single container mass of 30 tons. The lifting equipment is equipped with a 200kW generator-motor integrated machine. When the container is lifted to a height of 25 meters, it can store approximately 2.04kWh of potential energy. The state of charge of gravity energy storage is calculated in real time through container position sensors and weight sensors. The equivalent state of charge is defined as the ratio of the current lifting height to the maximum lifting height. The supercapacitor energy storage module has a rated capacity of 100F, a rated voltage of 1500V, and a maximum charging and discharging power of 200kW. The voltage, temperature, and state of charge of each individual cell are monitored in real time.

[0024] Specifically, the DC bus module is divided into a high-voltage DC bus and a low-voltage DC bus; the high-voltage DC bus is connected to the photovoltaic power generation module, the wind power generation module, and the gravity energy storage module; the low-voltage DC bus is connected to the hoisting mechanism inverter and auxiliary facilities; the high-voltage DC bus and the low-voltage DC bus are bidirectionally connected through a bidirectional DC / DC converter.

[0025] In an optional embodiment, the rated voltage of the high-voltage DC bus is 1500V, and the operating voltage range is 1350-1500V; the rated voltage of the low-voltage DC bus is 620V, and the operating voltage range is 540-750V; the high-voltage and low-voltage buses are connected in parallel by two 500kW bidirectional DC / DC converters.

[0026] Furthermore, the energy management module is used to monitor the operating status of each module in real time, adopts a prediction-correction two-layer architecture to formulate energy allocation strategies, and realizes coordinated control of each module. The prediction-correction two-layer architecture is as follows: the prediction layer is based on time-of-use electricity prices, long-term load forecasts, initial energy storage capacity, and operation plans, and uses mixed integer linear programming to formulate energy storage charging and supply plans, equipment start-up and shutdown plans, and grid interaction plans for the next 24 hours, with an update cycle of 6-24 hours; the correction layer is based on real-time electricity prices, medium-term load forecasts, and actual energy storage capacity, and uses model predictive control to correct the plans for the next 4-8 hours, with an update cycle of 15-60 minutes. The energy management module uses an industrial-grade PLC controller, which communicates with each module through a bus, with a sampling period of 1000ms and a control period of 1000ms.

[0027] Based on the teachings of the above embodiments, other aspects of the present invention also propose a port multi-energy coordinated control method based on peak-valley electricity prices and load forecasting, including: S1: Collect meteorological data, operational data, historical data, and external data. After preprocessing the collected data, input it into the LSTM-Attention prediction model to perform long-term, medium-term, and short-term load forecasts. The predicted output values ​​include photovoltaic power generation, wind power generation, port electricity consumption, and grid power supply.

[0028] It should be noted that the meteorological data collected includes irradiance, temperature, humidity, wind speed, wind direction, cloud cover, and precipitation; the operational data collected includes ship arrival times, container throughput, and equipment operating status; the historical data collected includes load power, photovoltaic output, wind power output, and energy storage capacity; and the external data collected includes time-of-use electricity prices, real-time electricity prices, grid frequency, tidal times, and holiday information. The collected data is processed according to the Laida criterion to remove gross errors, and then normalized and feature-values ​​are extracted. The specific process is as follows: Figure 2 As shown.

[0029] Furthermore, the LSTM-Attention model includes a two-layer LSTM network for extracting time series features and an Attention layer for calculating weight coefficients at different time steps. This model can map features to predicted values. The data is input into the LSTM-Attention prediction model, and the model parameters are shown in Table 1.

[0030] Table 1. LSTM-Attention Model Parameters In an optional embodiment, the LSTM-Attention model has 64 LSTM units, a Dropout value of 0.2, an optimizer of Adam, a loss function of MSE, a learning rate of 0.001, a batch size of 32, and 600 training epochs.

[0031] Furthermore, the long-term forecast duration is 24-48 hours with a time resolution of 1 hour, the medium-term forecast duration is 4-8 hours with a time resolution of 15 minutes, and the short-term forecast duration is 15-60 minutes with a time resolution of 5 minutes.

[0032] S2: Based on long-term load forecasts, combined with time-of-use electricity price information and the initial state of charge of energy storage modules, and with the goal of minimizing overall operating costs, mixed-integer linear programming is used to solve the forecast scheduling plan within the forecast time period. The forecast scheduling plan includes energy storage charging and discharging plan, photovoltaic / wind power generation plan and grid interaction plan.

[0033] Specifically, the objective function for optimizing the objective includes: Operatecost=Σ[Cdown(t)×Pdown(t)×Δt]+Σ[Cup(t)×Pup(t)×Δt]; Where Operatecost is the operating cost, Cdown(t) is the power price for disconnecting from the network at the t-th scheduling time, Pdown(t) is the power for disconnecting from the network at the t-th scheduling time, Cup(t) is the power price for connecting to the network at the t-th scheduling time, Pup(t) is the power for connecting to the network at the t-th scheduling time, and Δt is the scheduling time step.

[0034] Furthermore, the constraints on the optimization objectives include power balance constraints, energy storage capacity constraints, charging power constraints, discharging power constraints, photovoltaic power output constraints, and wind power output constraints; among which, the power balance constraints are: PPV(t)+PWT(t)+PGdischarge(t)+PSCdischarge(t)+Pgrid(t)=Pload(t)+PGcharge(t)+PSCcharge(t); Energy storage capacity constraint: SOCmin ≤ SOC(t) ≤ SOCmax; Charging power constraint: Pcharge_min≤Pcharge(t)≤Pcharge_max; Discharge power constraint: Pdischarge_min≤Pdischarge(t)≤Pdischarge_max; Photovoltaic output constraint: 0 ≤ PPV(t) ≤ PPV_max(t); Wind power output constraint: 0 ≤ PWT(t) ≤ PWT_max(t); Wherein, PPV(t) is the photovoltaic output, PWT(t) is the wind power output, PGdischange(t) is the gravity energy storage power supply, PSCdischange(t) is the supercapacitor power supply, Pgrid(t) is the grid power, Pload(t) is the port load, PGchange(t) is the gravity energy storage charging power, PSCchange(t) is the supercapacitor charging power, SOC(t) is the energy storage capacity, SOCmin is the minimum energy storage capacity, SOCmax is the maximum energy storage capacity, Pchange(t) is the charging power, Pchange_min is the minimum charging power, Pchange_max is the maximum charging power, Pdischange(t) is the power supply, Pdischange_min is the minimum power supply, Pdischange_max is the maximum power supply, PPV_max(t) is the maximum photovoltaic output, and PWT_max(t) is the maximum wind power output.

[0035] In an optional embodiment, as shown in Table 2, the time-of-use electricity price is divided into off-peak hours (23:00-5:00), flat hours (5:00-10:00, 12:00-16:00, 20:00-23:00) and peak hours (10:00-12:00, 16:00-20:00); in the initial state of charge of the energy storage module, the SOC of gravity energy storage is 50%, and the SOC of supercapacitor is 70%.

[0036] Table 2. Predicted Output Data Table Furthermore, the synergistic charging and discharging strategy of solar power, wind power, and energy storage includes: Collaborative charging strategy: When PPV(t) + PWT(t) ≥ Pload(t), the surplus power Psurplus(t) = PPV(t) + PWT(t) - Pload(t) is calculated, and the supercapacitor is charged first, with the supercapacitor charging power PSCcharge(t) = min[Psurplus(t), PSCcharge_max]; if there is still surplus power or the supercapacitor's state of charge has reached its upper limit, gravity energy storage is charged, with the gravity energy storage charging power PGcharge(t) = min[Psurplus(t) - PSCcharge(t), PGcharge_max]; when the gravity energy storage's state of charge has also reached its upper limit, the surplus power is fed back to the grid. Coordinated Discharge Strategy: When PPV(t) + PWT(t) < Pload(t), calculate the power gap Pshort(t) = Pload(t) - PPV(t) - PWT(t). First, the supercapacitor supplies power, and the supercapacitor discharge power PSCdischarge(t) = min[Pshort(t), PSCdischarge_max]; if there is still a power gap or the state of charge of the supercapacitor has reached the lower limit, then the gravity energy storage supplies power, and the gravity energy storage discharge power PGdischarge(t) = min[Pshort(t) - PSCdischarge(t), PGdischarge_max]; when the state of charge of the gravity energy storage also reaches the lower limit, the remaining power gap is filled by the power grid; Among them, PPV(t) is the photovoltaic output, PWT(t) is the wind power output, Pload(t) is the port load, Psurplus(t) is the surplus power, PSCchange(t) is the supercapacitor charging power, PSCcharge_max is the maximum supercapacitor charging power, PGdischange(t) is the gravity energy storage power supply, Pshort(t) is the power gap, PSCdischange(t) is the supercapacitor power supply, PSCdischarge_max is the maximum supercapacitor power supply, PGchange(t) is the gravity energy storage charging power, PGcharge_max is the maximum gravity energy storage charging power, Pdischange(t) is the power supply, and Pdischange_max is the maximum power supply.

[0037] In an optional embodiment, the typical daily 24-hour scheduling plan obtained by solving through mixed integer linear programming is shown in Table 3; during the valley electricity period (23:00 - 5:00), there is no photovoltaic output, the gravity energy storage uses low-cost grid electricity to charge 130 kW, and the supercapacitor does not participate temporarily; during the flat electricity period (5:00 - 10:00), the photovoltaic output gradually increases, the gravity energy storage discharges 100 kW to supplement the power gap, and the supercapacitor smooths the photovoltaic output fluctuations; during the peak electricity period (10:00 - 12:00), the photovoltaic and wind power outputs basically meet the load demand, the gravity energy storage discharges a small amount of 10 kW, and the supercapacitor responds to the instantaneous power fluctuations; during the period of large photovoltaic generation (12:00 - 16:00), the photovoltaic output significantly exceeds the load demand, the gravity energy storage charges a large amount of 250 kW, and 80 kW of surplus electricity is fed back to the power grid; during the evening peak period (16:00 - 20:00), the photovoltaic output decreases but the load is high, the gravity energy storage discharges 160 kW to supplement the gap, and the supercapacitor continuously responds to the fluctuations; during the flat electricity period (20:00 - 23:00), the renewable energy output is insufficient, the gravity energy storage discharges 100 kW, and the power grid supplements the remaining gap of 130 kW.

[0038] Table 3. Typical Daily 24-Hour Scheduling Plan S3: Based on the medium-term load forecast and forecast scheduling plan, combined with real-time electricity price, actual energy storage capacity and photovoltaic / wind power forecast errors, model predictive control is used to correct the scheduling plan within the forecast period.

[0039] Specifically, the corrections include corrections for photovoltaic output, wind power output, and load: Photovoltaic output correction value: PPV_corrected(t) = PPV_pred(t) + αPV × ΔPPV; Wind power output correction value: PWT_corrected(t) = PWT_pred(t) + αWT × ΔPWT; Load correction value: Pload_corrected(t) = Pload_pred(t) + αload × ΔPload; Wherein, PPV_pred(t) is the predicted photovoltaic output, ΔPPV is the deviation between the actual photovoltaic output and the predicted value, and αPV is the photovoltaic smoothing coefficient; PWT_pred(t) is the predicted wind power output, ΔPWT is the deviation between the actual wind power output and the predicted value, and αWT is the wind power smoothing coefficient; Pload_pred(t) is the predicted load, ΔPload is the deviation between the actual load and the predicted value, and αload is the load smoothing coefficient; Based on various correction values, the system energy storage charging and discharging plan and the grid interaction plan are continuously revised.

[0040] In an optional embodiment, during the peak power period at 10:00 on a certain day, the actual photovoltaic output is 80kW lower than the predicted value, and the actual load is 50kW higher than the predicted value. Taking αPV=0.4 and αload=0.4, the photovoltaic output correction value PPV_corrected(t)=350+0.4×(-80)=318kW, and the load correction value Pload_corrected(t)=450+0.4×50=470kW. Based on this, the gravity energy storage discharge power is corrected from 10kW to 40kW, and 22kW of power is supplemented from the grid, realizing real-time closed-loop correction of the power gap. The system control plan is adjusted according to the above correction values, and the comparison before and after the correction is shown in Table 4.

[0041] Table 4. Comparison of Scheduling Plans Before and After Revision In summary, this invention, through step S1, collects multi-source data including meteorological data, operational data, historical data, and external data. Combined with preprocessing and an LSTM-Attention model, it achieves load and output predictions at long, medium, and short time scales. This enables the system to obtain prediction results with temporal continuity and adaptive feature weights, providing high-confidence prior information support for subsequent scheduling. This process not only achieves quantitative modeling of port load fluctuations and the uncertainty of wind and solar power output but also effectively improves the foresight and interpretability of multi-energy system operation decisions, ultimately reducing the impact of prediction bias on system operation and improving overall scheduling accuracy. Step S2, based on long-term prediction results, introduces a time-of-use pricing mechanism and initial state constraints for energy storage, constructing a mixed-integer linear programming model with the objective of minimizing comprehensive operating costs. This achieves synergy between photovoltaic, wind, supercapacitor, and gravity energy storage. The process of optimizing the configuration of charging and discharging decisions and grid interaction unifies the modeling of energy supply and demand balance constraints and economic objectives. This not only achieves the global optimal allocation among multiple energy sources, but also strengthens the functional division of different energy storage systems over time through a hierarchical scheduling strategy that prioritizes supercapacitors and then gravity energy storage. This significantly reduces electricity costs and improves the renewable energy absorption rate while ensuring system stability. Step S3 introduces a model predictive control mechanism based on medium-term forecast results, real-time electricity prices, and actual operating deviations. This mechanism performs rolling corrections to the established scheduling plan and combines deviation correction models for photovoltaic, wind, and load power to achieve dynamic closed-loop control. This allows the system to continuously correct prediction errors and adapt to external disturbances during actual operation. This process effectively compensates for the shortcomings of traditional static scheduling in terms of real-time performance and robustness, thereby improving system operational stability, enhancing resistance to uncertainty, and ensuring the continuous optimization of the scheduling strategy.

[0042] This embodiment also provides a computer device applicable to the port multi-energy coordinated control method based on peak-valley electricity prices and load forecasting, including a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the port multi-energy coordinated control method based on peak-valley electricity prices and load forecasting as proposed in the above embodiment.

[0043] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0044] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the port multi-energy coordinated control method based on peak-valley electricity prices and load forecasting as proposed in the above embodiments.

[0045] The storage medium proposed in this embodiment and the data storage method proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0046] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A port multi-energy coordinated control system based on peak-valley electricity prices and load forecasting, characterized in that, include: Photovoltaic power generation modules are used to convert solar energy into electrical energy, which is then connected to a common DC bus via a DC converter. Wind power generation module, used to convert wind energy into electrical energy, which is then connected to a common DC bus via an AC / DC converter; Gravity energy storage module is used to store potential energy by using standard port containers as the energy storage medium. The container is lifted to a predetermined height by a lifting mechanism and the potential energy is released to generate electricity by a lowering mechanism, realizing the bidirectional conversion of potential energy and electrical energy. The supercapacitor energy storage module is used to provide instantaneous power support, smooth system power fluctuations, recover regenerative braking energy generated during crane operation, and connect to the common DC bus through a bidirectional DC converter; The DC bus module is used as an energy collection and distribution center to realize DC grid connection of multi-source energy; The energy management module is used to monitor the operating status of each module in real time, adopts a predictive-correction two-layer architecture to formulate energy allocation strategies, and realizes coordinated control of each module; The photovoltaic power generation module, the wind power generation module, the gravity energy storage module, and the supercapacitor energy storage module are all connected in parallel to the common DC bus of the DC bus module.

2. The port multi-energy coordinated control system based on peak-valley electricity prices and load forecasting as described in claim 1, characterized in that, The DC bus module is divided into a high-voltage DC bus and a low-voltage DC bus; the high-voltage DC bus is connected to the photovoltaic power generation module, the wind power generation module, and the gravity energy storage module; the low-voltage DC bus is connected to the hoisting mechanism inverter and auxiliary facilities; the high-voltage DC bus and the low-voltage DC bus are bidirectionally connected through a bidirectional DC / DC converter.

3. The port multi-energy coordinated control system based on peak-valley electricity prices and load forecasting as described in claim 2, characterized in that, The output power of the photovoltaic power generation module and the wind power generation module adaptively adjusts in accordance with the DC bus voltage, including: When the DC bus voltage is higher than the rated voltage, the output power is reduced to suppress the bus overvoltage. When the DC bus voltage is lower than the rated voltage, the maximum power is output to meet the load demand. When the sum of the output power of the photovoltaic power generation module and the wind power generation module is greater than the actual load demand, the surplus power is calculated and distributed to each energy storage module for charging in the order of priority of supercapacitor energy storage over gravity energy storage. When all energy storage modules reach their charging limit, the excess power is fed back to the grid.

4. The port multi-energy coordinated control system based on peak-valley electricity prices and load forecasting as described in claim 3, characterized in that, The DC-DC converter employs maximum power point tracking control, enabling the photovoltaic output power to adaptively adjust with the DC bus voltage.

5. The port multi-energy coordinated control system based on peak-valley electricity prices and load forecasting as described in claim 4, characterized in that, The wind power generation module is equipped with a maximum power point tracking controller, which dynamically adjusts the wind turbine speed and output power according to changes in wind speed.

6. A port multi-energy coordinated control method based on peak-valley electricity prices and load forecasting, based on the port multi-energy coordinated control system based on peak-valley electricity prices and load forecasting as described in any one of claims 1 to 5, characterized in that, include: S1: Collect meteorological data, operational data, historical data and external data. After preprocessing the collected data, input it into the LSTM-Attention prediction model to perform long-term, medium-term and short-term load forecasts. The predicted output values ​​include photovoltaic power generation, wind power generation, port electricity consumption and grid power supply. S2: Based on long-term load forecasts, combined with time-of-use electricity price information and the initial state of charge of energy storage modules, with the goal of minimizing overall operating costs, mixed integer linear programming is used to solve the forecast scheduling plan within the forecast time period, wherein the forecast scheduling plan includes energy storage charging and discharging plan, photovoltaic / wind power generation plan and grid interaction plan; S3: Based on the medium-term load prediction value and the predicted scheduling plan, combined with the real-time electricity price, the actual energy storage capacity, and the photovoltaic / wind power prediction error, use model predictive control to correct the scheduling plan within the prediction time.

7. The port multi-energy coordinated control method based on peak-valley electricity prices and load forecasting as described in claim 6, characterized in that, The constraint conditions of the optimization objective include power balance constraint, energy storage capacity constraint, charging power constraint, discharging power constraint, photovoltaic output constraint, and wind power output constraint; the objective function of the optimization objective includes: Operatecost=Σ[Cdown(t)×Pdown(t)×Δt]+Σ[Cup(t)×Pup(t)×Δt]; where, Operatecost is the operating cost, Cdown(t) is the power price of the power grid connection at the t-th scheduling moment, Pdown(t) is the power of the power grid connection at the t-th scheduling moment, Cup(t) is the power price of the power grid injection at the t-th scheduling moment, Pup(t) is the power of the power grid injection at the t-th scheduling moment, and Δt is the scheduling time step.

8. The port multi-energy coordinated control method based on peak-valley electricity prices and load forecasting as described in claim 7, characterized in that, The coordinated charging and discharging strategy of photovoltaic, wind power, and energy storage in step S2 includes: Coordinated charging strategy: When PPV(t)+PWT(t)≥Pload(t), calculate the surplus power Psurplus(t)=PPV(t)+PWT(t)-Pload(t), and preferentially charge the supercapacitor. The charging power of the supercapacitor PSCcharge(t)=min[Psurplus(t), PSCcharge_max]; if there is still surplus power or the state of charge of the supercapacitor has reached the upper limit, charge the gravity energy storage. The charging power of the gravity energy storage PGcharge(t)=min[Psurplus(t)-PSCcharge(t), PGcharge_max]; when the state of charge of the gravity energy storage also reaches the upper limit, the surplus power is fed back to the power grid; Coordinated discharging strategy: When PPV(t)+PWT(t)<Pload(t), calculate the power gap Pshort(t)=Pload(t)-PPV(t)-PWT(t), and preferentially supply power from the supercapacitor. The discharging power of the supercapacitor PSCdischarge(t)=min[Pshort(t), PSCdischarge_max]; if there is still a power gap or the state of charge of the supercapacitor has reached the lower limit, supply power from the gravity energy storage. The discharging power of the gravity energy storage PGdischarge(t)=min[Pshort(t)-PSCdischarge(t), PGdischarge_max]; when the state of charge of the gravity energy storage also reaches the lower limit, the remaining power gap is supplemented by the power grid; Wherein, PPV(t) is the photovoltaic output, PWT(t) is the wind power output, Pload(t) is the port load, Psurplus(t) is the surplus power, PSCchange(t) is the supercapacitor charging power, PSCcharge_max is the maximum supercapacitor charging power, PGdischange(t) is the gravity storage power supply, Pshort(t) is the power gap, PSCdischange(t) is the supercapacitor power supply, PSCdischarge_max is the maximum supercapacitor power supply, PGchange(t) is the gravity storage charging power, PGcharge_max is the maximum gravity storage charging power, Pdischange(t) is the power supply, and Pdischange_max is the maximum power supply.

9. The port multi-energy coordinated control method based on peak-valley electricity prices and load forecasting as described in claim 8, characterized in that, The corrections in step S3 include photovoltaic output correction, wind power output correction, and load correction; based on the correction values, the system energy storage charging and discharging plan and grid interaction plan are rolled over and corrected.

10. The port multi-energy coordinated control method based on peak-valley electricity prices and load forecasting as described in claim 9, characterized in that, It also includes, The photovoltaic output correction value is: PPV_corrected(t) = PPV_pred(t) + αPV × ΔPPV; The wind power output correction value is: PWT_corrected(t) = PWT_pred(t) + αWT × ΔPWT; The load correction value is: Pload_corrected(t) = Pload_pred(t) + αload × ΔPload; Wherein, PPV_pred(t) is the predicted photovoltaic output, ΔPPV is the deviation between the actual photovoltaic output and the predicted value, and αPV is the photovoltaic smoothing coefficient; PWT_pred(t) is the predicted wind power output, ΔPWT is the deviation between the actual wind power output and the predicted value, and αWT is the wind power smoothing coefficient; Pload_pred(t) is the predicted load, ΔPload is the deviation between the actual load and the predicted value, and αload is the load smoothing coefficient.

Images