Cooperative interaction intelligent regulation and control method for source network load storage resources

Abstract

The invention discloses a collaborative interaction intelligent regulation and control method for source network load storage resources, and relates to the technical field of resource regulation and control, and the method comprises the following steps: Sa1, obtaining the real-time fluctuation of a power grid load and the power grid flow distribution in a power grid system, and carrying out the prediction of the power flow distribution based on the real-time fluctuation of the power grid load and new energy power generation power prediction data; generating a supply-demand imbalance characteristic quantity of the power supply and the load; sa2, according to the supply and demand imbalance characteristic quantity, constructing a multi-objective optimization model containing participation of an energy storage system and an adjustable load; sa3, outputting a regulation and control instruction set by solving the multi-objective optimization model; and Sa4, carrying out coordination control on the source network load storage resources according to the regulation and control instruction set so as to realize dynamic balance. The method can solve the problem of how to regulate and control the new energy generation power according to the real-time fluctuation of the power grid load so as to solve the problem of mismatching between the power supply output and the load demand, thereby ensuring the stable power supply.

Description

Technical Field

[0001] The present invention relates to the technical field of resource regulation, and in particular to a collaborative interactive intelligent regulation method for source, grid, load and storage resources. Background Art

[0002] The coordinated regulation of source, grid, load and storage resources is achieved by integrating the multi-dimensional resources of power sources, power grids, loads and energy storage systems, utilizing advanced information communication and intelligent control technologies to optimize energy distribution and improve the operating efficiency and stability of the power system.

[0003] However, this method still faces the problem of how to effectively regulate the power generation of renewable energy according to the real-time fluctuations of the grid load. This is because renewable energy power generation is intermittent and uncertain, and the power output and load demand in the power system often do not match. If the power generation power cannot be accurately adjusted to adapt to the load changes, it may lead to unstable power supply or waste of resources, which puts higher requirements on the balance and reliability of the power system. The real-time fluctuations of the grid load regulate the power generation power of renewable energy to solve the mismatch between power output and load demand, and the changes in the grid current may cause voltage over-limit or instability in local areas. If the voltage support cannot be accurately monitored and reasonably allocated, it will affect the normal operation of the distribution network and the power supply quality. Summary of the Invention

[0004] The purpose of the present invention is to provide a collaborative interactive intelligent control method for source, grid, load and storage resources, which solves the mismatch between power output and load demand, and regulates the distribution network voltage level according to the grid flow distribution to solve the problem of local voltage exceeding the limit and instability.

[0005] The present invention solves the above technical problems through the following technical solutions, which include the following steps: Sa1. Obtain the real-time fluctuation of grid load and grid power flow distribution within the power grid system, and generate the supply and demand imbalance characteristic of power source and load based on the real-time fluctuation of grid load and the forecast data of renewable energy power generation; Sa2. Constructing a multi-objective optimization model involving an energy storage system and adjustable loads based on the supply-demand imbalance characteristic quantity; Sa3. Outputting a control instruction set by solving the multi-objective optimization model; Sa4. Coordinate and control the source, grid, load and storage resources according to the control instruction set to achieve dynamic balance.

[0006] Preferably, generating the power supply and load supply and demand imbalance characteristic quantity based on the real-time fluctuation of the power grid load includes the following steps: Obtain the real-time grid load fluctuation value P_load and the predicted value of renewable energy power generation power P_gen; Calculate the initial power difference ΔP_init = |P_load P_gen| as the basic unbalance; Adjust the feature size according to the following formula: ΔP_final = α * ΔP_init, where α is the system adjustment coefficient (0 < α ≤ 1), which represents the intensity of the supply and demand imbalance feature adjustment; If ΔP_final exceeds the preset maximum imbalance threshold P_threshold, ΔP_final is limited to P_threshold.

[0007] Preferably, the adjustment method for the supply-demand imbalance characteristic quantity is further optimized: Collect historical load data and calculate the average load level P_avg; The degree of change of the current load state relative to the historical average level is calculated through the deviation ratio λ = |P_load P_avg| / P_avg; Update the system regulation coefficient α = α * (1 + γ * λ) according to the following formula, where γ is the proportional gain factor; Make sure the new system adjustment coefficient α is within the range [α_min, α_max].

[0008] Preferably, logic for dynamically evaluating supply and demand balance characteristics is added: Define the short-term sliding average time window length T_window and calculate the average fluctuation amplitude A within the window; The recent volatility severity R is determined by the formula R = (1 / N) * Σ(|P_load_i P_avg_i|), i ∈ {1, ..., N}, and N is the number of samples; According to the current supply and demand status, if ΔP_final / R ≥ δ, it means that the imbalance is aggravated and the control response level needs to be increased; δ is the sensitivity coefficient; The corresponding safety redundancy parameters are set according to the response level, and the final imbalance characteristic quantity is regenerated.

[0009] Preferably, the energy storage device state constraint is considered in the characteristic quantity adjustment: Detect the remaining capacity E_remain and the maximum supported output / charge capacity E_lim of the energy storage system; When E_remain / E_lim ≤ β (β is the safe operation ratio, usually 1 ≤ β ≤ 0.2), reduce the weight of the characteristic variable to limit excessive control intensity; The correction factor θ is introduced to calculate the adjusted ΔP_final = θ * (E_remain / E_lim) * ΔP_init; θ represents the energy storage impact coefficient; The adjustment result ΔP_final is used in the subsequent control algorithm design to ensure stable operation of energy storage.

[0010] Preferably, analyzing the power grid flow distribution includes the following steps: Sb1. Based on the analysis of the power flow distribution results, determine the areas where the voltage exceeds the limit and is unstable in the distribution network. Sb2. Construct an optimization control model for source, grid, load and storage resources to generate a preliminary control strategy; Sb3. Verify and adjust the preliminary control strategy to form the final control instructions; Sb4. Coordinate and dispatch source, grid, load and storage resources according to the final control instructions to improve the distribution network voltage level.

[0011] Preferably, the determining of voltage exceeding limit and unstable areas in the distribution network based on the power grid power flow distribution analysis result further includes: Obtain the current and power flow distribution parameters $ P_i $ and $ Q_i $ in the target area; Use the following formula to determine whether voltage control is needed: $\sqrt{P_i^2 + Q_i^2} > U_i \cdot Z_i$, where $ U_i $ is the voltage at node i and $ Z_i $ is the equivalent impedance; If the formula result is true, then the voltage instability exists in the corresponding area of the mark; Otherwise the recorded area is a stable area.

[0012] Preferably, the determining of voltage exceeding limit and unstable areas in the distribution network based on the power grid power flow distribution analysis result further includes: Collect line load rate data $L_j$ and voltage fluctuation data $V_j$ in the target area; Calculate the voltage sensitivity factor $F_v$: $F_v = (\alpha \cdot V_j \beta \cdot L_j) / C_s$, where $\alpha, \beta$ are weight coefficients and $C_s$ is a normalization constant; When $ F_v < 1 $, the corresponding area is divided into the key focus area; Prioritize and mark the instability level based on the number of voltage anomalies in the key focus area.

[0013] Preferably, the analysis result based on the power grid flow distribution further includes: Extract key connection information from the distribution network topology and organize the load distribution matrix $G_d$; Calculate the load imbalance: $B_u = (|R_{max}| |R_{min}|) / |G_d|$, where $R_{max}$ and $R_{min} $ represent the ratio of maximum to minimum load; When the regional load ratio exceeds $ 5\cdot B_u $, the partition is recorded as a potential risk point; The boundary of the high load area is updated by superimposing the flow direction data of multiple branches.

[0014] Preferably, determining the voltage exceeding the limit and unstable region further comprises the following process: Get the historical data $V_{hist,i}$ of all partitions to be verified; Set the warning threshold $ Th_r = U_N \cdot (1+\delta) $, where $ \delta $ is the safety margin factor and $ U_N $ is the rated voltage level; The following discriminant is used to check the level of out-of-bounds: $ E_v = |U_i U_N| / Th_r $, when $ E_v > 2 $, it is marked as a severely out-of-bounds partition; The marking results are passed to the next stage regulation module to generate a preliminary control strategy.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. The present invention can solve the problem of how to regulate the power generation power of renewable energy according to the real-time fluctuation of the grid load to solve the problem of mismatch between power output and load demand, thereby ensuring stable power supply.

[0016] 2. The present invention can solve the problem of how to regulate the distribution network voltage level according to the power grid flow distribution to solve the problem of local voltage exceeding the limit and instability, thereby ensuring the normal operation of the distribution network and good power supply quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a flow chart of embodiment 1 of the present invention; Figure 2 This is a flow chart of embodiment 2 of the present invention. DETAILED DESCRIPTION

[0018] Example 1 This embodiment provides a technical solution: a method for intelligently controlling source, grid, load, and storage resources through collaborative interaction. The method regulates the power generated by renewable energy sources based on real-time grid load fluctuations to resolve the mismatch between power output and load demand, including the following steps: Sa1. Obtain the real-time fluctuation of grid load and grid power flow distribution within the power grid system, and generate the supply and demand imbalance characteristic of power source and load based on the real-time fluctuation of grid load and the forecast data of renewable energy power generation; Sa2. Constructing a multi-objective optimization model involving an energy storage system and adjustable loads based on the supply-demand imbalance characteristic quantity; Sa3. Outputting a control instruction set by solving the multi-objective optimization model; Sa4. Coordinate and control the source, grid, load and storage resources according to the control instruction set to achieve dynamic balance.

[0019] The method of generating a power supply and load supply and demand imbalance characteristic value based on real-time power grid load fluctuations includes the following steps: Obtain the real-time grid load fluctuation value P_load and the predicted value of renewable energy power generation power P_gen; Calculate the initial power difference ΔP_init = |P_load P_gen| as the basic unbalance; Adjust the feature size according to the following formula: ΔP_final = α * ΔP_init, where α is the system adjustment coefficient (0 < α ≤ 1), which represents the intensity of the supply and demand imbalance feature adjustment; If ΔP_final exceeds the preset maximum imbalance threshold P_threshold, ΔP_final is limited to P_threshold.

[0020] Further optimized the adjustment method for the supply and demand imbalance characteristic quantity: Collect historical load data and calculate the average load level P_avg; The degree of change of the current load state relative to the historical average level is calculated through the deviation ratio λ = |P_load P_avg| / P_avg; Update the system regulation coefficient α = α * (1 + γ * λ) according to the following formula, where γ is the proportional gain factor; Make sure the new system adjustment coefficient α is within the range [α_min, α_max].

[0021] Added logic for dynamically evaluating supply and demand balance features: Define the short-term sliding average time window length T_window and calculate the average fluctuation amplitude A within the window; The recent volatility severity R is determined by the formula R = (1 / N) * Σ(|P_load_i P_avg_i|), i ∈ {1, ..., N}, and N is the number of samples; According to the current supply and demand status, if ΔP_final / R ≥ δ, it means that the imbalance is aggravated and the control response level needs to be increased; δ is the sensitivity coefficient; The corresponding safety redundancy parameters are set according to the response level, and the final imbalance characteristic quantity is regenerated.

[0022] Preferably, the energy storage device state constraint is considered in the characteristic quantity adjustment: Detect the remaining capacity E_remain and the maximum supported output / charge capacity E_lim of the energy storage system; When E_remain / E_lim ≤ β (β is the safe operation ratio, usually 1 ≤ β ≤ 0.2), reduce the weight of the characteristic variable to limit excessive control intensity; The correction factor θ is introduced to calculate the adjusted ΔP_final = θ * (E_remain / E_lim) * ΔP_init; θ represents the energy storage impact coefficient; The adjustment result ΔP_final is used in the subsequent control algorithm design to ensure stable operation of energy storage.

[0023] This example first defines the key steps involved: first, generating characteristic quantities of supply and demand imbalances between power sources and loads; second, constructing a multi-objective optimization model; third, solving the optimization model to output a set of control instructions; and fourth, coordinating and controlling source, grid, load, and storage resources based on the control instructions. This method achieves precise matching of power sources and loads in dynamic scenarios by comprehensively analyzing various factors in the power system, such as real-time grid load fluctuations and renewable energy generation power forecasts.

[0024] Specifically, in the first step, a characteristic measure of the supply-demand imbalance between power sources and loads is generated based on real-time grid load fluctuations and forecasted renewable energy generation data. This step collects and organizes real-time grid operating data to provide an accurate information foundation. Specifically, advanced measuring instruments, monitoring platforms, and distributed sensor technology collect current grid load data in real time. This data is then combined with weather forecast models or historical statistical data to predict future renewable energy generation trends, such as the projected power output of wind turbines in the coming hours or the potential efficiency of solar photovoltaic panels. Using mathematical statistical analysis, this input information is converted into variable representations for further modeling. This generates a key reference indicator, the potential deviation between supply (actual and forecasted power output at the power source) and demand (the immediate demand level at the user end), which is referred to as the characteristic measure of the supply-demand imbalance.

[0025] After obtaining the characteristic quantity of supply-demand imbalance, the second step begins. Based on this characteristic quantity, a multi-objective optimization model is established that integrates the energy storage system and a load factor with flexible adjustment capabilities. To ensure optimal overall system efficiency, this optimization model considers objectives such as reducing energy waste, lowering operating costs, and maintaining power quality. This requires the use of complex algorithms, such as genetic algorithms or particle swarm optimization, to achieve a balance under various constraints. For example, energy storage systems can play an important role in temporarily absorbing excess power or releasing stored power to alleviate supply shortages. Adjustable loads, on the other hand, are specific devices that can temporarily reduce power consumption or postpone work schedules to help mitigate peak demand. For example, within an industrial park, there are numerous remotely controlled production facilities. When the external power grid exceeds its load limit, pre-emptive measures can be implemented to suspend the operation of certain essential facilities during different periods.

[0026] The third step then involves using advanced mathematical tools or specialized software to rapidly solve the complex multi-objective optimization model. This results in a set of control solutions, known as a control instruction set. Each specific instruction corresponds to a list of actions to be performed at a specific time interval, such as when a pumped-storage power station should pump hydropower and convert it into potential energy for later use, or whether several smart air conditioning units in a given area should simultaneously activate energy-saving mode to reduce overall consumption to an optimal range.

[0027] Finally, the precise instructions obtained above are used to effectively control the entire resource integration system, encompassing the source, network node interconnection, end-user load distribution, and the participation of various energy storage devices, to achieve dynamic balance maintenance. This means that all participants act according to pre-set rules and continuously monitor and correct feedback to adapt to environmental changes and ensure that the goals are achieved.

[0028] For example, in one embodiment, imagine that one evening, a sudden drop in large-scale photovoltaic power generation leads to a power shortage in the regional distribution network. This process first identifies this widening supply-demand gap, quickly assesses its extent, and then invokes a pre-designed policy framework to organize response actions. For example, this involves discharging backup batteries to replenish the power shortage while also issuing recommended power restrictions to contracted customers. These coordinated actions promptly resolve potential issues, restore normalcy, and effectively support the robust and secure operation of the entire power grid.

[0029] Next, the method of generating a power supply and load supply and demand imbalance characteristic value based on real-time power grid load fluctuation according to the present invention includes the following steps: First, we list the individual steps. The first step is to obtain the real-time grid load fluctuation value \( P_{\text{load}} \) and the predicted renewable energy power generation value \( P_{\text{gen}} \). The second step is to calculate the initial power difference \( \Delta P_{\text{init}} = |P_{\text{load}} - P_{\text{gen}}| \) based on these values and use this as the basic imbalance value. The third step is to adjust the characteristic value using the formula \( \Delta P_{\text{final}} = \alpha * \Delta P_{\text{init}} \), where the adjustment coefficient \( \alpha \) satisfies \( 0 <\alpha \leq 1 \) and represents the intensity of the adjustment of the supply and demand imbalance characteristic. The last step is to determine if the final adjusted imbalance amount \( \Delta P_{\text{final}} \) exceeds the preset maximum imbalance threshold \( P_{\text{threshold}} \), then \( \Delta P_{\text{final}} \) will be forced to be equal to \( P_{\text{threshold}} \).

[0030] The first step is to determine current load demand and power generation output using actual monitoring or simulation data. Grid load fluctuations reflect the dynamic power consumption of consumers within the power system over time. Meanwhile, power generation forecasts represent the projected output of renewable energy sources such as wind and photovoltaic power. These fluctuate significantly due to factors like weather, requiring real-time estimation to match demand fluctuations. Comparing these two provides first-hand data for assessing the power balance.

[0031] The second step directly compares the two parameters to obtain the difference, using absolute calculations to ensure the result is always positive. This initial power difference is used to quantify the unfilled power shortage or surplus within the system during a specific transient state. The key to this step is accurately capturing the imbalance. Therefore, it determines the prerequisites for the subsequent regulation logic.

[0032] The third step introduces an adjustment coefficient, \( \alpha \), to further correct the initially calculated imbalance value to better suit the control requirements of specific scenarios. The parameter \( \alpha \) ranges from near zero (reflecting a very slight imbalance) to no more than one, with a default value of 0.9 being the preferred value. This formula is designed to provide varying flexibility and response accuracy under different system conditions, avoiding situations where overly aggressive control can lead to ineffective regulation. Specifically, low values of \( \alpha \) prioritize suppressing the risk of large adjustments, while higher values increase sensitivity, allowing even small differences to be quickly fed back to the control unit for processing.

[0033] The fourth step aims to set a protection limit to ensure that the entire system does not operate beyond the safe load capacity. When it is detected that the maximum imbalance threshold is exceeded, it means that an extreme situation may have been triggered and it needs to be temporarily fixed at the limit state before manual intervention to prevent it from deteriorating into irreversible damage. For example, during a period of strong wind but low load, if it exceeds the normal allowable range, the output must be forcibly constrained to ensure the stability of the overall structure. In one embodiment, the problem of local overpower supply caused by a sudden increase in solar irradiance during the day in a certain area of the wind, solar and storage integrated scenario is specifically considered. At this time, if it is deduced from the above steps that the initial imbalance reaches about twice the carrying capacity of the power grid, then according to the rules, it will be suppressed to within the threshold level line and the subsequent trend will be observed before deciding whether to release the excess energy into the energy storage pool or other absorption paths.

[0034] Next, the present invention further optimizes the adjustment method for the supply-demand imbalance characteristic quantity: first, the steps of collecting historical load data, calculating the deviation ratio, updating the system adjustment coefficient, and limiting the range of the system adjustment coefficient are listed.

[0035] In the first step, historical load data is collected over a period of time to obtain the average load level \(P_{\text{avg}}\). This refers to a representative value obtained by summarizing and analyzing the power load values within a specific time interval in the past, which is used as a reference benchmark.

[0036] The second step involves quantifying the degree of change in the current load state based on the difference between the current load and the historical average. This is calculated using the formula \(\lambda = |P_{\text{load}} - P_{\text{avg}}| / P_{\text{avg}}\), where \(P_{\text{load}}\) represents the current actual load and \(\lambda\) accurately reflects the degree of load change relative to the average. Typically, the value of \(\lambda\) ranges from zero to a reasonable upper limit, depending on the application scenario. The optimal value is determined based on actual operating experience to achieve stability.

[0037] Then, in the third step, the original coefficient is dynamically adjusted using \(\alpha = \alpha * (1 + \gamma * \lambda)\). Here, \(\alpha\) indicates the initial control coefficient or the previous set value in the system, while \(\gamma\) is the proportional gain factor used as a proportional coefficient for adjusting the degree of influence. This coefficient can be selected from a reasonable value range based on the different characteristics of the system (for example, when \(\lambda\) fluctuates greatly, a relatively small value such as 0.05 can be selected to avoid excessive interference). The optimal setting is determined by experimental verification, such as around 0.1.

[0038] Finally, after adjusting the coefficients, it is necessary to check whether the new \(\alpha\) remains within a specified range \([\alpha_{\text{min}}, \alpha_{\text{max}}]\) to avoid adverse consequences caused by excessively high or low values. The boundary values set depend on the allowable operating limits of the specific power network and equipment.

[0039] For example, in one embodiment, assume that historical data for a grid region indicates a normal daily load level of 5 MW (\(P_{\text{avg}}=5MW\)). If the load measurement result this morning suddenly rises to 8 MW, the above method can be used to quickly assess and respond. Specifically, the deviation calculation shows \(\lambda=|8-5| / 5=0.6\), and then the coefficient \(\alpha\) is updated according to the set scaling factor (\(\gamma=0.1\)) to become the original value multiplied by 1.06, and then checked to see if it meets the established range. This helps to adapt to power changes in real time and ensure the safe and economic operation of the power system.

[0040] Next, the logic of the present invention that adds the dynamic evaluation of the supply and demand balance feature is described:

[0041] A short-term sliding average window length, T_window, is defined to select a recent period of power data for analysis. This operation, through a sliding average, filters out noise and the influence of short-term random fluctuations, ensuring data representativeness. The time window range can be determined by the actual application scenario. For example, in power systems, time periods ranging from minutes to hours may be optimal to accommodate the varying speeds of load characteristics.

[0042] The average fluctuation amplitude within the window is then quantified using the formula R = (1 / N) * Σ(|P_load_i - P_avg_i|), where i ∈ {1, ..., N}. Here, R is an indicator of the severity of load fluctuations within the window; N represents the number of sampled data points; P_load_i represents the actual load power value at each moment in the window; and P_avg_i is the average load power calculated over the T_window period at the same moment. This formula is designed to measure the stability of recent load fluctuations, providing a reliable benchmark for predicting future trends. This average deviation form provides a smoother evaluation metric.

[0043] After completing the above assessment, further assessment is needed based on the current overall supply and demand relationship. If ΔP_final / R ≥ δ, the current system imbalance is considered to be increasing, and the control response level must be increased to match supply and demand. Here, ΔP_final reflects the real-time net difference (supply exceeds demand or vice versa), and δ is a manually set sensitivity coefficient, usually determined based on experience or test optimization to reflect the system's sensitivity to imbalances. If the sensitivity is too high, it may trigger an overly aggressive control response, while if it is too low, it may lead to delayed adjustments.

[0044] Finally, based on the comprehensive assessment results, the corresponding redundancy parameters are adjusted according to different response levels. This is then used to regenerate the final imbalance signature, ensuring sufficient robustness and sensitivity when all links work together. In one embodiment, assuming a T_window of 30 minutes, the most recent 20 data samples are collected for evaluation. The calculation reveals that R is large, and the ratio of the absolute value of ΔP_final to R exceeds a preset threshold δ. At this point, safety redundancy is increased to a higher level, such as adding an additional 10% backup power capacity. This information is then fed into the next decision-making process to achieve the optimal scheduling strategy.

[0045] Next, we will describe how the present invention considers energy storage device state constraints when adjusting characteristic quantities. First, we will list and explain each step in sequence. These steps include: first, detecting the remaining capacity and maximum supported output / input of the energy storage system; second, setting weight reduction conditions based on the energy storage system's safe operation ratio; third, introducing a correction factor to calculate the final power adjustment; and fourth, applying these adjustment results to the control algorithm.

[0046] Step one involves determining the real-time data of the energy storage device through testing. This step requires obtaining two key parameters: the energy storage system's remaining charge (E_remain) and its maximum allowable output or charging capacity (E_lim). Both quantities are positive real-valued, and their values depend on the specific energy storage technology. This testing process can be accomplished by analyzing data from sensors within the energy storage battery.

[0047] Step two involves reducing the amount of adjustment to reduce system burden in specific situations. When the ratio of the current capacity to the upper capacity limit is less than or equal to a predetermined value, β, the limiting operation is triggered. β is defined as the safe operating ratio threshold allowed by the system. In this scheme, it is typically in the range of 0.2 to 1, with β = 0.2 being ideally selected. This is because a lower ratio means the system is operating close to the minimum pressure limit or the maximum load range, and increasing the adjustment too much may lead to irreversible changes in the system state.

[0048] Next, we move on to step three. Here, we introduce a variable called the correction factor θ, and perform the final adjustment calculation according to the formula: ΔP_final = θ * (E_remain / E_lim) * ΔP_init. Here, θ represents a coefficient that measures the impact of the overall performance of the energy storage system on global optimization. The reasonable range for θ is 0 to 1, with a recommended set point of 0.5. This allows the initial calculated power change to be appropriately scaled to meet the aforementioned safety requirements. This formula is designed to provide flexible guidance for scheduling within the entire model, taking into account the actual capacity of the energy storage units.

[0049] For the last part, which is the fourth step, it is to use the adjusted final value obtained in the previous step as the basic input factor for other related logical judgments to continue to develop in depth.

[0050] Example 2 This embodiment provides a technical solution: a method for collaboratively interacting and intelligently regulating source, grid, load, and storage resources, for regulating the distribution network voltage level based on the power grid flow distribution to address local voltage over-limit and instability, including the following steps: Analyzing the power flow distribution of the power grid includes the following steps: Sb1. Determine the voltage exceeding the limit and unstable areas in the distribution network based on the analysis of the power flow distribution results; Sb2. Construct an optimization control model for source, grid, load and storage resources to generate a preliminary control strategy; Sb3. Verify and adjust the preliminary control strategy to form the final control instructions; Sb4. Coordinate and dispatch source, grid, load and storage resources according to the final control instructions to improve the distribution network voltage level.

[0051] The determining of voltage exceeding limit and unstable areas in the distribution network based on the power grid flow distribution analysis result further includes: Obtain the current and power flow distribution parameters $ P_i $ and $ Q_i $ in the target area; Use the following formula to determine whether voltage control is needed: $\sqrt{P_i^2 + Q_i^2} > U_i \cdot Z_i$, where $ U_i $ is the voltage at node i and $ Z_i $ is the equivalent impedance; If the formula result is true, then the voltage instability exists in the corresponding area of the mark; Otherwise the recorded area is a stable area.

[0052] The determining of voltage exceeding limit and unstable areas in the distribution network based on the power grid flow distribution analysis result further includes: Collect line load rate data $L_j$ and voltage fluctuation data $V_j$ in the target area; Calculate the voltage sensitivity factor $F_v$: $F_v = (\alpha \cdot V_j \beta \cdot L_j) / C_s$, where $\alpha, \beta$ are weight coefficients and $C_s$ is a normalization constant; When $ F_v < 1 $, the corresponding area is divided into the key focus area; Prioritize and mark the instability level based on the number of voltage anomalies in the key focus area.

[0053] The analysis results based on the power grid flow distribution further include: Extract key connection information from the distribution network topology and organize the load distribution matrix $G_d$; Calculate the load imbalance: $B_u = (|R_{max}| |R_{min}|) / |G_d|$, where $R_{max}$ and $R_{min} $ represent the ratio of maximum to minimum load; When the regional load ratio exceeds $ 5\cdot B_u $, the partition is recorded as a potential risk point; The boundary of the high load area is updated by superimposing the flow direction data of multiple branches.

[0054] Determining the voltage exceeding the limit and unstable areas further includes the following process: Get the historical data $V_{hist,i}$ of all partitions to be verified; Set the warning threshold $ Th_r = U_N \cdot (1+\delta) $, where $ \delta $ is the safety margin factor and $ U_N $ is the rated voltage level; The following discriminant is used to check the level of out-of-bounds: $ E_v = |U_i U_N| / Th_r $, when $ E_v > 2 $, it is marked as a severely out-of-bounds partition; The marking results are passed to the next stage regulation module to generate a preliminary control strategy.

[0055] In this embodiment, the specific areas requiring regulation are first determined based on the results of the grid power flow distribution analysis. The key to this step is to assess the current load distribution and line conditions within the grid and identify areas experiencing excessively high or low voltages. For example, during peak hours or when the load is unbalanced, a specific transformer may output a voltage exceeding the safe range. Accurately identifying these over-limit and unstable conditions is crucial for the successful implementation of subsequent regulation strategies.

[0056] After clarifying the above issues, an optimization and control model that comprehensively considers power sources, power grids, loads, and energy storage resources is constructed to generate a preliminary control strategy. Specifically, this step includes collecting various information related to sources, grids, loads, and storage, such as the real-time power generation capacity of generators, fluctuations in user-end power demand, the actual output level of distributed renewable energy power generation equipment, and the current remaining capacity of energy storage equipment. These factors are then linked together using mathematical modeling. These factors are quantified and input into specific intelligent optimization algorithms, such as linear programming algorithms, dynamic optimization algorithms, or heuristic search algorithms, to develop a preliminary plan, that is, to form the initial operational guidance plan.

[0057] The initially formed control strategy is then further refined and improved through a rigorous verification process, ultimately resulting in a reliable execution instruction system. The verification process primarily considers two key areas: safety checks to prevent improper configurations from causing new system risks or damaging equipment and facilities; and economic efficiency assessments to ensure that the entire system operates at the lowest possible cost while maintaining high-quality service levels. If, during the verification phase, it is discovered that certain pre-set solutions do not adequately address overall benefits, the process will be revised and optimized based on the specific circumstances, ensuring that each decision is both responsive to changing needs and sustainable in the long term.

[0058] Finally, according to the instructions officially issued after the adjustment is completed, all source, grid, load and storage elements are coordinated and directed to take synchronous actions to improve the voltage quality at each point in the distribution network to achieve the purpose of operating within the predetermined standard range.

[0059] The above are only preferred embodiments of the present invention and are merely illustrative, not restrictive, of the present invention. Those skilled in the art will appreciate that many changes, modifications, and even equivalents may be made to the embodiments within the spirit and scope of the claims, all of which fall within the scope of protection of the present invention.

Claims

1. A collaborative interactive intelligent control method for source, grid, load and storage resources, characterized in that: The following steps are involved: Sa1. Obtain the real-time fluctuation of grid load and grid power flow distribution within the power grid system, and generate the supply and demand imbalance characteristic of power source and load based on the real-time fluctuation of grid load and the forecast data of renewable energy power generation; Sa2. Constructing a multi-objective optimization model involving an energy storage system and adjustable loads based on the supply-demand imbalance characteristic quantity; Sa3. Outputting a control instruction set by solving the multi-objective optimization model; Sa4. Coordinate and control the source, grid, load and storage resources according to the control instruction set to achieve dynamic balance.

2. The method for collaborative interactive intelligent control of source, grid, load and storage resources according to claim 1, characterized in that: The method of generating a power supply and load supply and demand imbalance characteristic value based on real-time power grid load fluctuations includes the following steps: Obtain the real-time grid load fluctuation value P_load and the predicted value of renewable energy power generation power P_gen; Calculate the initial power difference ΔP_init = |P_load P_gen| as the basic unbalance; Adjust the feature size according to the following formula: ΔP_final = α * ΔP_init, where α is the system adjustment coefficient (0 < α ≤ 1), which represents the intensity of the supply and demand imbalance feature adjustment; If ΔP_final exceeds the preset maximum imbalance threshold P_threshold, ΔP_final is limited to P_threshold.

3. The method for collaborative interactive intelligent control of source, grid, load and storage resources according to claim 2, characterized in that: Further optimized the adjustment method for the supply and demand imbalance characteristic quantity: Collect historical load data and calculate the average load level P_avg; The degree of change of the current load state relative to the historical average level is calculated through the deviation ratio λ = |P_load P_avg| / P_avg; Update the system regulation coefficient α = α * (1 + γ * λ) according to the following formula, where γ is the proportional gain factor; Make sure the new system adjustment coefficient α is within the range [α_min, α_max].

4. The method for collaborative interactive intelligent control of source, grid, load and storage resources according to claim 3, characterized in that: Added logic for dynamically evaluating supply and demand balance features: Define the short-term sliding average time window length T_window and calculate the average fluctuation amplitude A within the window; The recent volatility severity R is determined by the formula R = (1 / N) * Σ(|P_load_i P_avg_i|), i ∈ {1, ..., N}, and N is the number of samples; According to the current supply and demand status, if ΔP_final / R ≥ δ, it means that the imbalance is aggravated and the control response level needs to be increased; δ is the sensitivity coefficient; The corresponding safety redundancy parameters are set according to the response level, and the final imbalance characteristic quantity is regenerated.

5. The method for collaborative interactive intelligent control of source, grid, load and storage resources according to claim 4, characterized in that: Consider the energy storage device state constraints in the characteristic quantity adjustment: Detect the remaining capacity E_remain and the maximum supported output / charge capacity E_lim of the energy storage system; When E_remain / E_lim ≤ β (β is the safe operation ratio, usually 1 ≤ β ≤ 0.2), reduce the weight of the characteristic variable to limit excessive control intensity; The correction factor θ is introduced to calculate the adjusted ΔP_final = θ * (E_remain / E_lim) * ΔP_init; θ represents the energy storage impact coefficient; The adjustment result ΔP_final is used in the subsequent control algorithm design to ensure stable operation of energy storage.

6. The method for collaborative interactive intelligent control of source, grid, load and storage resources according to claim 1, characterized in that: Analyzing the power flow distribution of the power grid includes the following steps: Sb1. Based on the analysis of the power flow distribution results, determine the areas where the voltage exceeds the limit and is unstable in the distribution network. Sb2. Construct an optimization control model for source, grid, load and storage resources to generate a preliminary control strategy; Sb3. Verify and adjust the preliminary control strategy to form the final control instructions; Sb4. Coordinate and dispatch source, grid, load and storage resources according to the final control instructions to improve the distribution network voltage level.

7. The method for collaborative interactive intelligent control of source, grid, load and storage resources according to claim 6, characterized in that: The determining of voltage exceeding limit and unstable areas in the distribution network based on the power grid flow distribution analysis result further includes: Obtain the current and power flow distribution parameters $ P_i $ and $ Q_i $ in the target area; Use the following formula to determine whether voltage control is needed: $\sqrt{P_i^2 + Q_i^2} > U_i \cdot Z_i$, where $ U_i $ is the voltage at node i and $ Z_i $ is the equivalent impedance; If the formula result is true, then the voltage instability exists in the corresponding area of the mark; Otherwise the recorded area is a stable area.

8. The method for collaborative interactive intelligent control of source, grid, load and storage resources according to claim 1, characterized in that: The determining of voltage exceeding limit and unstable areas in the distribution network based on the power grid flow distribution analysis result further includes: Collect line load rate data $L_j$ and voltage fluctuation data $V_j$ in the target area; Calculate the voltage sensitivity factor $F_v$: $F_v = (\alpha \cdot V_j \beta \cdot L_j) / C_s$, where $\alpha, \beta$ are weight coefficients and $C_s$ is a normalization constant; When $ F_v < 1 $, the corresponding area is divided into the key focus area; Prioritize and mark the instability level based on the number of voltage anomalies in the key focus area.

9. The method for collaborative interactive intelligent control of source, grid, load and storage resources according to claim 3, characterized in that: The analysis results based on the power grid flow distribution further include: Extract key connection information from the distribution network topology and organize the load distribution matrix $G_d$; Calculate the load imbalance: $B_u = (|R_{max}| |R_{min}|) / |G_d|$, where $R_{max} $ and $R_{min} $ represent the ratio of maximum to minimum load; When the regional load ratio exceeds $ 5\cdot B_u $, the partition is recorded as a potential risk point; The boundary of the high load area is updated by superimposing the flow direction data of multiple branches.

10. The method for collaborative interactive intelligent control of source, grid, load and storage resources according to claim 4, characterized in that: Determining the areas of voltage violations and instability further includes the following processes: Get the historical data $V_{hist,i}$ of all partitions to be verified; Set the warning threshold $ Th_r = U_N \cdot (1+\delta) $, where $ \delta $ is the safety margin factor and $ U_N $ is the rated voltage level; The following discriminant is used to check the level of out-of-bounds: $ E_v = |U_i U_N| / Th_r $, when $ E_v > 2 $, it is marked as a severely out-of-bounds partition; The marking results are passed to the next stage regulation module to generate a preliminary control strategy.

Images