A low-carbon-oriented smart park multi-energy complementary and collaborative scheduling method

Abstract

This invention relates to the field of power system operation and control technology, specifically to a multi-energy complementary and coordinated scheduling method for smart parks aimed at low-carbon goals. The method includes: collecting multi-dimensional state data of the smart park and performing standardized preprocessing; constructing a multi-objective coordinated scheduling cost function and physical constraints; performing a rolling solution on the multi-objective coordinated scheduling cost function and physical constraints to obtain the optimal control sequence; and issuing the first instruction in the optimal control sequence to the energy storage converter in the smart park's power grid for execution, thus achieving closed-loop control. The solution of this invention significantly extends the service life of energy storage devices while achieving dynamic low-carbon scheduling.

Description

Technical Field

[0001] This invention relates to the field of power system operation and control technology. More specifically, this invention relates to a smart park multi-energy complementary and coordinated scheduling method oriented towards low-carbon goals. Background Technology

[0002] As a core unit of urban energy consumption, the low-carbon transformation of smart parks has become an inevitable trend in industry development. In actual operation, smart parks typically include various power source devices such as distributed photovoltaic power generation systems, electrochemical energy storage systems, electric vehicle charging piles, and conventional office loads.

[0003] Existing smart park energy dispatching methods primarily focus on economic optimization, namely, formulating strategies based on time-of-use (time-of-use) electricity pricing (peak-valley pricing) to charge or purchase electricity during low-price periods and discharge or reduce electricity purchases during high-price periods. However, existing technologies have significant limitations in addressing the goal of "low carbon": On the one hand, carbon-electricity coupling quantification suffers from a lag. Existing technologies typically use fixed carbon emission factors to calculate carbon emissions. However, the actual carbon emission intensity of the power grid changes dynamically over time (e.g., carbon emissions are lower during peak solar power generation at midday and higher during peak evening hours when thermal power is the main regulator). Static indicators cannot capture these millisecond- or minute-level carbon flow fluctuations, leading to situations where, although the industrial park saves on electricity costs, it may purchase large amounts of electricity during periods of high carbon emissions from the grid, resulting in "pseudo-low-carbon" operation.

[0004] On the other hand, the lifespan degradation of energy storage devices is often overlooked. In order to respond to fluctuations in photovoltaic power and complex low-carbon directives, energy storage batteries often require frequent power regulation. Existing greedy algorithms do not adequately consider the battery's health status, leading to accelerated aging due to frequent charge-discharge switching and drastic power fluctuations, increasing the replacement cost over the entire lifespan.

[0005] Therefore, there is an urgent need for a collaborative scheduling method that can quantify the carbon-electric coupling relationship in real time and effectively protect the lifespan of energy storage devices while achieving low-carbon scheduling. Summary of the Invention

[0006] The purpose of this invention is to propose a multi-energy complementary and coordinated scheduling method for smart parks aimed at low carbon goals, in order to solve the problem that existing technologies cannot achieve a balance between low carbon and the lifespan of energy storage equipment, thus leading to pseudo-low carbon operation and accelerated aging of energy storage equipment; to this end, this invention provides a solution in one aspect.

[0007] This invention provides a smart park multi-energy complementary and coordinated scheduling method for low-carbon goals, comprising: Collect and preprocess multidimensional status data of the smart park. The multidimensional status data includes at least real-time power generation, total load demand, real-time time-of-use electricity price of the power grid, real-time dynamic carbon emission intensity of the power grid, and state of charge of energy storage batteries. A multi-objective collaborative scheduling cost function is constructed, which includes a power purchase potential energy driving term based on the carbon-electric coupling potential energy index and a battery action damping term; the carbon-electric coupling potential energy index is positively correlated with the real-time time-of-use electricity price and the real-time dynamic carbon emission intensity. Physical constraints are constructed, which are obtained from real-time power generation, total load demand, and state of charge of energy storage batteries. The multi-objective collaborative scheduling cost function and physical constraints are solved in a rolling manner to obtain the optimal control sequence. The first instruction in the optimal control sequence is then sent to the energy storage converter of the power grid in the smart park for execution, thereby realizing closed-loop control.

[0008] The above scheme constructs a multi-objective collaborative scheduling cost function that includes a carbon-electric coupling potential energy index driving term and a battery action damping term. On the one hand, it uses the potential energy index to deeply integrate economic and environmental goals, achieving a keen perception of dynamic carbon emissions. On the other hand, it introduces a battery action damping term to give the system physical inertia, which effectively smooths out the power fluctuations of the energy storage battery while achieving dynamic low-carbon scheduling, and significantly extends the service life of the energy storage equipment.

[0009] Optionally, the formula for calculating the carbon electrocoupling potential energy index is: ; In the formula, Let be the carbon-electric coupling potential energy index at time t; Let be the real-time electricity price of the power grid at time t; The benchmark electricity price constant; Let be the real-time carbon emission intensity of the power grid at time t; The historical average carbon emission intensity constant of the regional power grid; and These are the weighting coefficients for economic potential and low-carbon potential, respectively. This is the carbon sensitivity factor.

[0010] The above scheme overcomes the limitations of traditional linear weighting methods by constructing a nonlinear carbon-electric coupling potential energy index that includes logarithmic terms (for electricity prices) and power terms (for carbon emissions). By utilizing the nonlinear growth characteristics of the power function, the algorithm becomes highly sensitive to periods of high carbon emissions, enabling the park to automatically form a "high potential energy repulsion zone" during high-carbon periods on the power grid, forcibly reducing electricity purchases and thus avoiding "pseudo-low-carbon" operation, achieving true dynamic low-carbon optimization.

[0011] Optionally, the benchmark electricity price constant is taken as the historical lowest off-peak electricity price of the smart park, and the historical average carbon emission intensity constant is used as a benchmark reference line for carbon intensity; the carbon sensitivity factor The value of is greater than 1.

[0012] Optionally, the battery action damping term for: ; in, To predict the first in the time domain The charging and discharging power of the energy storage battery at any given moment; To predict the first in the time domain Battery power at any given moment; This is the battery action damping coefficient. This is the scheduling time step.

[0013] The above scheme defines a battery action damping term based on the square of the power change rate (first derivative) to impose a secondary penalty on the drastic changes in the power of the energy storage battery, so that the battery charge and discharge curve changes from "aggressive adjustment" to "smooth adjustment", simulating the inertial characteristics of the physical system. This smooth control effectively reduces the heat accumulation and side reactions inside the battery, and significantly reduces the battery capacity decay rate compared with the greedy response strategy of the existing technology.

[0014] Optionally, the multi-objective cooperative scheduling cost function The calculation formula is: ; In the formula, To predict the length of the time domain; To predict the first in the time domain The carbon-electric coupling potential energy index at a given moment; To predict the first in the time domain Power purchased by the power grid at any given moment; This is the battery action damping term.

[0015] The above scheme endows the scheduling system with a kind of "physical inertia" through the design of a multi-objective collaborative scheduling cost function, which makes the battery charge and discharge curve present a smooth shape, effectively reducing the heat accumulation and side reactions inside the battery. While ensuring low-carbon benefits, it significantly slows down the battery capacity decay and extends the service life of the equipment.

[0016] Optionally, the preset physical constraints include power balance constraints and battery capacity constraints; The power balance constraint satisfies: ; The battery capacity constraint satisfies: ; in, Let be the total load demand power at time t+k. Let be the real-time power generation at time t+k. For the target power purchase at time t+k, Let be the charging or discharging power of the energy storage battery at time t+k. Let be the state of charge of the energy storage battery at time t+k. and These are the minimum and maximum limits for the state of charge, respectively.

[0017] The above scheme ensures that the control commands generated by the scheduling algorithm are physically executable by clearly defining power balance constraints and battery capacity constraints. This not only guarantees the supply and demand balance of power sources within the park, but also strictly limits the state of charge of the batteries within a safe range (preventing overcharging and over-discharging), thus ensuring the safe and stable operation of the entire microgrid system.

[0018] Optionally, the battery state of charge for: ; In the formula, For the effective capacity of the battery, To predict the first in the time domain The state of charge of the battery at a given moment. To predict the first in the time domain The charging and discharging power of the energy storage battery at any given moment. This is the scheduling time step.

[0019] Optionally, the specific steps of the rolling solution are as follows: Within the prediction time domain, the goal is to minimize the multi-objective cooperative scheduling cost function. A quadratic programming solver is used to solve the problem and obtain the optimal control sequence for N time steps. The optimal control sequence includes the energy storage battery charging and discharging power sequence. The first energy storage battery charge / discharge power command in the energy storage battery charge / discharge power sequence is selected and sent to the energy storage converter for control and execution.

[0020] Optionally, the implementation of closed-loop control includes: After executing the instructions for the current moment, wait to enter the next scheduling moment; The new multidimensional state data of the smart park is repeatedly collected, and the carbon-electric coupling potential energy index is recalculated based on the new multidimensional state data, thus restarting the rolling solution process.

[0021] Optionally, real-time power generation, total load demand, real-time time-of-use electricity price of the grid, and real-time dynamic carbon emission intensity of the grid can be obtained through photovoltaic inverters, regional grid dispatch interfaces, and smart building systems; and data cleaning, interpolation, and time alignment processing can be performed on photovoltaic power generation, load demand, and grid-side data respectively.

[0022] The beneficial effects of this invention are as follows: The solution of this invention achieves keen perception of high carbon emission periods through a nonlinear potential energy field model and smooths the charge and discharge curves using a physical inertial damping mechanism, thereby significantly extending the service life of energy storage equipment while realizing dynamic low-carbon scheduling. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating a smart park multi-energy complementary and coordinated scheduling method for low-carbon goals according to an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the time-series variation of the carbon electrocoupling potential energy index according to an embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the smoothness comparison of energy storage battery scheduling commands according to an embodiment of the present invention; Figure 4 This is a schematic diagram illustrating the comparison of power purchase response under different scheduling strategies according to embodiments of the present invention. Detailed Implementation

[0024] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0025] This invention proposes a multi-energy complementary and coordinated scheduling method for smart parks aimed at low-carbon goals. By constructing a dynamic potential energy field and introducing physical damping, it aims to solve the problems of carbon-electric coupling quantification lag and shortened lifespan of energy storage devices due to high-frequency power fluctuations in existing technologies.

[0026] The following example illustrates a smart park that includes a distributed photovoltaic power generation system, an electrochemical energy storage system, and conventional loads.

[0027] Specifically, such as Figure 1 As shown in this embodiment, a multi-energy complementary and coordinated scheduling method for smart parks aimed at low-carbon goals includes the following steps: Step S1: Collect multi-dimensional status data of the smart park and perform preprocessing.

[0028] In this embodiment, the first step is to acquire multi-dimensional status data of the smart park's real-time operation. Specifically, multi-dimensional status data is collected synchronously in real time through photovoltaic inverters, intelligent building systems (including smart meters and battery management systems) (BMS), and regional power grid dispatch interfaces.

[0029] The collected multidimensional state data includes at least the following: SourceLoad Data: Collects real-time power generation data from distributed photovoltaic systems via photovoltaic inverters. The total load demand of the smart park is collected through the intelligent building system. .

[0030] Grid data: Real-time time-of-use electricity prices are obtained from the regional power grid dispatch interface. and real-time dynamic carbon emission intensity .

[0031] Energy storage data: Reading the state of charge of energy storage batteries through smart building systems. Effective battery capacity .

[0032] In this embodiment, after acquiring the collected multidimensional state data, data cleaning, interpolation completion, and time alignment are performed on each dimension of the state data to eliminate the effects of data noise and timestamp asynchrony.

[0033] The time alignment process described above unifies and normalizes the time resolution of all dimensional state data into a scheduling time step.

[0034] For example, assume the scheduling step size The time interval is 15 minutes (0.25 hours). At a certain time t, the following multidimensional state data was collected: real-time power generation. Total load demand power At this time, the real-time time-of-use electricity price of the power grid Yuan / kWh (during peak electricity price period), while real-time dynamic carbon emission intensity (This indicates that thermal power accounts for a relatively high proportion of the power grid at this time), the current state of charge of the battery. The percentage is 60%. After cleaning and aligning the above multidimensional state data, preprocessed multidimensional state data is obtained.

[0035] The above-mentioned real-time acquisition and preprocessing of multi-source heterogeneous data eliminated data noise and timing deviations, providing a high-quality data foundation for subsequent accurate calculations.

[0036] Step S2: Construct the carbon-electric coupling potential energy index.

[0037] To address the problem that traditional static carbon emission factors cannot capture the dynamic changes in carbon flow in the power grid, this invention introduces the concept of "potential field" and constructs a carbon-electric coupling potential energy index to regard the external power grid as a potential field source. When electricity prices are high and carbon emissions are high, a "high potential energy repulsion zone" is formed, driving smart parks to reduce electricity purchases; conversely, a "low potential energy attraction zone" is formed.

[0038] Specifically, the carbon-electric coupling potential energy index The calculation formula is as follows: ; in, Let be the carbon-electric coupling potential energy index at time t; Let be the real-time time-of-use electricity price of the power grid at time t; As the benchmark electricity price constant, Let be the real-time dynamic carbon emission intensity of the power grid at time t. The historical average carbon emission intensity constant of the regional power grid; and These are the first and second weighting coefficients, respectively. ; This is the carbon sensitivity factor.

[0039] The aforementioned benchmark electricity price constant is taken as the historical lowest off-peak electricity price in the region, such as... The historical average carbon emission intensity constant can be set to a value of yuan / kWh. The values ​​of the first and second weighting coefficients mentioned above can be 0.6 and 0.4, respectively.

[0040] The carbon sensitivity factor must be greater than 1 to amplify the effects of high carbon emissions nonlinearly. In this embodiment, the carbon sensitivity factor... The preferred value is between 1.5 and 2.5. For example, when setting... At that time, if the real-time carbon emission intensity It is the average value. of The resulting penalty potential energy will be amplified to times. times; if It is the average value The penalty potential energy will be amplified to times. This nonlinear design makes the system highly sensitive to periods of high carbon emissions, forcing it to immediately reduce electricity purchases when the grid's carbon intensity increases slightly, thus avoiding "pseudo-low-carbon" operation.

[0041] The economic potential energy in the above formula is in logarithmic form, simulating the law of diminishing marginal returns. When the real-time time-of-use electricity price is low, small fluctuations in the real-time time-of-use electricity price have a significant impact on dispatch strategies; when the real-time time-of-use electricity price is already high, further increases have a milder inhibitory effect on the willingness to purchase electricity.

[0042] For example, when the first weighting coefficient The second weighting coefficient is 0.6. The carbon sensitivity factor is 0.4. Benchmark electricity price constant Yuan / kWh, historical average carbon emission intensity , , At this time, the carbon-electric coupling potential energy index .

[0043] As can be seen from the above examples, if it is the midday photovoltaic power generation period, the real-time time-of-use electricity price is low and the carbon emission intensity is also low, then the carbon-electric coupling potential energy index is also low.

[0044] It is evident that the carbon-electric coupling potential energy index during the evening peak forms an extremely high potential energy barrier compared to the carbon-electric coupling potential energy index during the midday period, which will strongly inhibit the system from purchasing electricity from the grid.

[0045] like Figure 2 As shown, the solid line (the carbon-electric coupling potential energy index of this invention) shows a nonlinear steep rise during the evening peak (6 pm to 11 pm). At this time, the superposition of high carbon emission intensity and high electricity price forms the electricity purchase repulsion zone, that is, the high potential energy repulsion zone.

[0046] In the above embodiments, the carbon-electric coupling potential energy index constructed by the nonlinear formula can sensitively amplify high carbon emission and high electricity price signals, construct a "carbon barrier", and force the scheduling system to prioritize the use of clean energy, thus achieving true dynamic low-carbon optimization.

[0047] Step S3: Construct the multi-objective cooperative scheduling cost function and physical constraints.

[0048] Specifically, after obtaining the carbon-electric coupling potential energy index, it is necessary to decide whether the battery should be charged or discharged, and by how much. Therefore, this step constructs a multi-objective cooperative scheduling cost function. Furthermore, a battery action damping term based on physical inertia was introduced to protect the battery.

[0049] Among them, the multi-objective cooperative scheduling cost function The calculation formula is: ; in, To predict the length of the time domain; To predict the carbon-electric coupling potential energy exponent at time t+k in the time domain; To predict the power purchased by the power grid at time t+k in the time domain; This is to predict the battery action damping term at time t+k in the time domain.

[0050] The above battery action damping term The specific expression is as follows: ; in, To predict the charging and discharging power of the energy storage battery at time t+k in the time domain; To predict the first in the time domain Battery power at any given moment; For scheduling time steps (e.g.) minute); This represents the battery's action damping coefficient.

[0051] The above prediction time domain length This corresponds to the next 4 hours, meaning that starting from the current t-th moment, the next 4 hours are divided into 16 time steps.

[0052] From the above formula, we can see that the battery action damping term A second penalty was applied to the rate of change of battery power (i.e., the first derivative).

[0053] For example, suppose (per unit time), battery action damping coefficient So, the battery power at the previous moment There are two scheduling schemes: Option A (Aggressive Adjustment): The battery power suddenly increases to [value] at time t. Power change The value of the battery action damping term is .

[0054] Option B (Smooth Adjustment): The battery power gradually increases to [a certain value] at the current moment. Power change The value of the battery action damping term is .

[0055] From the two scheduling schemes mentioned above, it can be seen that scheme A generates 16 times the battery action damping term as scheme B. Therefore, during the optimization process, in order to improve the multi-objective cooperative scheduling cost function... To minimize this, the algorithm will tend to choose a smooth curve similar to scheme B, avoiding the sudden stop-start operation of scheme A.

[0056] Figure 3 This is a schematic diagram of the charging and discharging power of an energy storage battery. In this diagram, the solid line (the solution of this invention) has a significantly smoother and more continuous waveform compared to the dashed line (the greedy response of the prior art). In this case, the solution of this invention has a smoother power change compared to the frequent fluctuations of the prior art, which can extend the life of the energy storage battery.

[0057] When solving the above cost function, the following physical constraints must be satisfied: The power balance constraint satisfies the following formula: ; in, Let be the total load demand power at time t+k. Let be the real-time power generation at time t+k. For the target power purchase at time t+k, Let be the charging or discharging power of the energy storage battery at time t+k; where is the target power purchase. When the energy storage battery discharges During charging .

[0058] Battery capacity and status update constraints: ;in, Let be the state of charge of the energy storage battery at time t+k. and These are the minimum and maximum limits for the state of charge, respectively.

[0059] The dynamic update equation for the battery state of charge is as follows: Optionally, the battery state of charge for: ; In the formula, For the effective capacity of the battery, To predict the first in the time domain The state of charge of the battery at a given moment. To predict the first in the time domain The charging and discharging power of the energy storage battery at any given moment. This is the scheduling time step.

[0060] In this embodiment, Set as , Set as This is to prevent overcharging and over-discharging.

[0061] In the above embodiments, by introducing a damping term based on the square of the power change rate into the MPC model, the system is given "physical inertia", which effectively smooths out the power fluctuations of the energy storage battery and prevents frequent oscillations caused by responding to small potential energy fluctuations, thereby significantly extending the battery's service life.

[0062] Step S4: Perform rolling solution on the multi-objective cooperative scheduling cost function and physical constraints to achieve closed-loop control.

[0063] In this embodiment, a quadratic programming solver is used to perform rolling solutions on the multi-objective cooperative scheduling cost function and preset physical constraints to obtain the optimal control sequence.

[0064] Specifically, a quadratic programming (QP) solver is used to solve the optimization problem constructed above. The solver will output the future... The optimal control sequence for each time step is determined, and based on the rolling optimization principle of MPC, the first instruction in the optimal control sequence is selected. The command is then sent to the energy storage converter for execution.

[0065] Since quadratic programming solvers are existing technology, their specific solution process will not be described in detail here.

[0066] For example, the solver outputs the future Optimal control sequence at each time step At this point, the first instruction in the sequence is selected. It is then distributed to the power grid's power storage converter (PCS) within the smart park for execution.

[0067] Furthermore, the specific process of closed-loop control in this embodiment is as follows: After completing the current time step After receiving the instruction, proceed to the next scheduling time. At this point, the latest multi-dimensional status data is collected again, and steps S1 to S4 are repeated. This closed-loop rolling mechanism can correct prediction errors in a timely manner and cope with the uncertainties of photovoltaics and loads.

[0068] like Figure 4 As shown, the solid line represents the power purchased by the power grid under the strategy of this invention. During the evening peak high-carbon period, it is significantly lower than that of the prior art (dashed line). The shaded part is the power purchase reduced by the active carbon avoidance of this invention.

[0069] Through the aforementioned closed-loop rolling mechanism, this embodiment can not only cope with the uncertainties of photovoltaics and loads, but also capture the low window of the grid carbon intensity in real time through the carbon-electric coupling potential energy index for charging and discharging during the peak window of carbon intensity, thereby reducing the operating cost of the park and significantly reducing indirect carbon emissions.

[0070] The solution of this invention, through rolling solution and closed-loop control mechanism, enables the system to continuously track the real-time status changes of the smart park, ensuring the real-time performance and robustness of the scheduling strategy, and ensuring the implementation of low-carbon and economic goals.

[0071] In the description of this specification, "multiple" means at least two, such as two, three or more, etc., unless otherwise expressly and specifically defined.

[0072] While various embodiments of the invention have been shown and described in this specification, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention.

Claims

1. A smart park multi-energy complementary and coordinated scheduling method for low-carbon goals, characterized in that, include: Collect and preprocess multidimensional status data of the smart park. The multidimensional status data includes at least real-time power generation, total load demand, real-time time-of-use electricity price of the power grid, real-time dynamic carbon emission intensity of the power grid, and state of charge of energy storage batteries. A multi-objective collaborative scheduling cost function is constructed, which includes a power purchase potential energy driving term based on the carbon-electric coupling potential energy index and a battery action damping term; the carbon-electric coupling potential energy index is positively correlated with the real-time time-of-use electricity price and the real-time dynamic carbon emission intensity. Physical constraints are constructed, which are obtained from real-time power generation, total load demand, and state of charge of energy storage batteries. The multi-objective collaborative scheduling cost function and physical constraints are solved in a rolling manner to obtain the optimal control sequence. The first instruction in the optimal control sequence is then sent to the energy storage converter of the power grid in the smart park for execution, thereby realizing closed-loop control.

2. The method for multi-energy complementary and coordinated scheduling in smart parks oriented towards low-carbon goals, as described in claim 1, is characterized in that... The formula for calculating the carbon electrocoupling potential energy index is as follows: ； In the formula, Let be the carbon-electric coupling potential energy index at time t; Let be the real-time electricity price of the power grid at time t; The benchmark electricity price constant; Let be the real-time carbon emission intensity of the power grid at time t; The historical average carbon emission intensity constant of the regional power grid; and These are the weighting coefficients for economic potential and low-carbon potential, respectively. This is the carbon sensitivity factor.

3. The method for multi-energy complementary and coordinated scheduling in smart parks oriented towards low-carbon goals, as described in claim 2, is characterized in that... The benchmark electricity price constant is taken as the historical lowest off-peak electricity price of the smart park, and the historical average carbon emission intensity constant serves as a benchmark reference line for carbon intensity; the carbon sensitivity factor The value of is greater than 1.

4. The method for multi-energy complementary and coordinated scheduling in smart parks oriented towards low-carbon goals according to claim 1, characterized in that, The battery action damping term for: ； in, To predict the first in the time domain The charging and discharging power of the energy storage battery at any given moment; To predict the first in the time domain Battery power at any given moment; This is the battery action damping coefficient. This is the scheduling time step.

5. The method for multi-energy complementary and coordinated scheduling in smart parks oriented towards low-carbon goals, as described in claim 4, is characterized in that... The multi-objective cooperative scheduling cost function The calculation formula is: ； In the formula, To predict the length of the time domain; To predict the first in the time domain The carbon-electric coupling potential energy index at a given moment; To predict the first in the time domain Power purchased by the power grid at any given moment; This is the battery action damping term.

6. The method for multi-energy complementary and coordinated scheduling in smart parks oriented towards low-carbon goals, as described in claim 1, is characterized in that... The physical constraints include power balance constraints and battery capacity constraints. The power balance constraint satisfies: ； The battery capacity constraint satisfies: ; in, Let be the total load demand power at time t+k. Let be the real-time power generation at time t+k. For the target power purchase at time t+k, Let be the charging or discharging power of the energy storage battery at time t+k. Let be the state of charge of the energy storage battery at time t+k. and These are the minimum and maximum limits for the state of charge, respectively.

7. A smart park multi-energy complementary and coordinated scheduling method for low-carbon goals as described in claim 6, characterized in that, The state of charge of the energy storage battery for: ； In the formula, For the effective capacity of the battery, To predict the first in the time domain The state of charge of the battery at a given moment. To predict the first in the time domain The charging and discharging power of the energy storage battery at any given moment. This is the scheduling time step.

8. The method for multi-energy complementary and coordinated scheduling in smart parks oriented towards low-carbon goals, as described in claim 1, is characterized in that... The specific steps of the rolling solution are as follows: In the prediction time domain, with the goal of minimizing the multi-objective cooperative scheduling cost function, a quadratic programming solver is used to solve the problem and obtain the optimal control sequence for N time steps. The optimal control sequence includes the energy storage battery charging and discharging power sequence. The first energy storage battery charge / discharge power command in the energy storage battery charge / discharge power sequence is selected and sent to the energy storage converter for control and execution.

9. A smart park multi-energy complementary and coordinated scheduling method for low-carbon goals as described in claim 1, characterized in that, The implementation of closed-loop control includes: After executing the instructions for the current moment, wait to enter the next scheduling moment; The new multidimensional state data of the smart park is repeatedly collected, and the carbon-electric coupling potential energy index is recalculated based on the new multidimensional state data, thus restarting the rolling solution process.

10. A smart park multi-energy complementary and coordinated scheduling method for low-carbon goals as described in claim 1, characterized in that, The process of collecting and preprocessing multi-dimensional status data of the smart park includes: Through photovoltaic inverters, regional power grid dispatch interfaces, and intelligent building systems, real-time power generation, total load demand, real-time time-of-use electricity price of the power grid, and real-time dynamic carbon emission intensity of the power grid are obtained; and data cleaning, interpolation, and time alignment are performed on photovoltaic power generation, load demand, and power grid side data respectively.

Images