Multi-mode adaptive power management method

Abstract

The invention discloses a multi-mode self-adaptive power supply management method, which comprises the following steps of: 1, acquiring multi-source power input data in real time, including at least two of wind power generation power, solar power generation power, a power grid power supply state and a standby generator state, and synchronously acquiring a charge state of an energy storage unit and load demand power; and step 2, based on the environmental parameter prediction model, combining with meteorological data and a historical power generation curve, generating power generation prediction values of wind energy and solar energy in a future preset time period, through implementation of the method, the global optimization problem of multi-energy cooperation is fundamentally solved, and based on adaptive strategy switching of the wind-solar power generation prediction value and the real-time load demand, the power generation capacity of the wind energy and the solar energy is improved. The power supply stability is greatly improved, the mode switching frequency is reduced through a dynamic charging proportion distribution and hybrid power supply priority calling mechanism, and the loss of power devices is remarkably reduced; and the strategy re-calculation function when the load / energy suddenly changes ensures that the response time of the system is prolonged.

Description

Technical Field

[0001] This invention relates to the field of power management technology, and in particular to a multi-mode adaptive power management method. Background Technology

[0002] With the rapid development of renewable energy technologies, distributed power generation systems such as wind and solar power are increasingly being used in microgrids. However, multi-energy coordinated management faces the following technical challenges:

[0003] The challenge of energy volatility: Wind and solar power generation is significantly affected by weather conditions, and the output power is random and intermittent, making it difficult for traditional single energy dispatch strategies to achieve stable power supply;

[0004] Multi-objective optimization conflict: It is necessary to simultaneously consider economic efficiency (such as the cost of electricity purchase under time-of-use pricing), reliability (continuous power supply to the load), and equipment lifespan (such as the safe range of energy storage SOC). Existing solutions often only optimize a single objective.

[0005] Insufficient dynamic response: When there are sudden load changes or energy failures, the fixed-rule scheduling system cannot quickly generate adaptive strategies, which can easily lead to over-discharge / overcharging of energy storage or power outages.

[0006] In summary, there is an urgent need for a power management method that integrates multi-source prediction, dynamic decision-making, and adaptive execution to address the balance between economy, reliability, and lifetime in multi-energy collaboration. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention provides a multi-mode adaptive power management method, aiming to solve the aforementioned problems.

[0008] In a first aspect, this application provides a multi-mode adaptive power management method, characterized by comprising the following steps:

[0009] Step 1: Collect multi-source power input data in real time, including at least two of the following: wind power generation, solar power generation, grid power supply status, and standby generator status, and simultaneously acquire the energy storage unit's state of charge and load demand power.

[0010] Step 2: Based on the environmental parameter prediction model, combined with meteorological data and historical power generation curves, generate predicted values ​​for wind and solar power generation within a preset future time period;

[0011] Step 3: Construct a dynamic decision matrix. Input parameters include: current energy input priority, energy storage SOC safety threshold, load power fluctuation range, time-of-use electricity price data, and the predicted power generation value.

[0012] Step 4: Output the optimal energy scheduling strategy based on the dynamic decision matrix, the strategy including:

[0013] When the total output power of renewable energy exceeds the load demand, the energy storage charging mode is executed, and the charging ratio of wind power and solar power to energy storage is dynamically allocated.

[0014] When the total output power of renewable energy is less than the load demand and the energy storage SOC is greater than the lower threshold, the hybrid power supply mode is activated, and the energy storage discharge, grid power purchase or backup generator power supply is called in the preset priority order.

[0015] When the total output power of renewable energy is less than the load demand and the energy storage SOC is less than or equal to the lower limit threshold, the system will be forced to switch to the grid / backup power dominant mode.

[0016] Step 5: The power allocation unit executes the energy dispatch strategy, adjusting the power allocation ratio of each energy channel in real time, and triggering strategy recalculation when load changes or energy fluctuations exceed thresholds. By constructing a closed-loop management framework of data acquisition, prediction, decision-making, execution, and re-triggering, the global optimization problem of multi-energy coordination is fundamentally solved. Adaptive strategy switching based on wind and solar power generation forecasts and real-time load demand greatly improves power supply stability. Dynamic charging ratio allocation and hybrid power supply priority calling mechanism reduce the number of mode switching times and significantly reduce power device losses. Furthermore, the strategy recalculation function during load / energy changes ensures improved system response time.

[0017] Furthermore, the dynamic decision matrix in step 3 is constructed in the following way:

[0018] Energy priority coefficients are defined as follows: Renewable energy > Energy storage > Power grid > Backup generators;

[0019] Set SOC hierarchical management thresholds: including emergency protection lower limit, high-efficiency charging range, and overcharge protection upper limit;

[0020] A multi-objective optimization function is established, with the optimization objectives being: minimizing grid purchase costs, maximizing renewable energy utilization, and extending energy storage lifespan. The priority of the objectives is dynamically adjusted through weighting factors.

[0021] Furthermore, the energy storage charging mode in step 4 specifically includes:

[0022] If the instantaneous power generation difference between wind and solar energy is greater than or equal to a set threshold, the power balancing algorithm will be activated to limit the charging current of highly volatile energy sources.

[0023] If the energy storage SOC is in the high-efficiency charging range, enable maximum power point tracking charging.

[0024] If the energy storage SOC is close to SOC_H, switch to trickle charging and start the excess energy discharge circuit.

[0025] Furthermore, the hybrid power supply mode in step 4 further includes:

[0026] During off-peak hours of time-of-use electricity pricing, priority is given to grid power supply, while during peak hours, priority is given to energy storage discharge.

[0027] When the standby generator is called, the generator output power is automatically adjusted to match the load power curve, so that it operates in the range of optimal fuel efficiency.

[0028] Furthermore, it also includes fault tolerance mechanisms:

[0029] Real-time monitoring of abnormal voltage / current fluctuations in each energy module;

[0030] When a single energy module fails, it is automatically removed from the power supply matrix, and the energy allocation strategy is recalculated.

[0031] If the energy storage unit fails, it will switch to a bypass mode where renewable energy directly supplies the load and the grid / generator provides supplementary power.

[0032] Furthermore, the environmental parameter prediction model in step 2 is optimized in the following way:

[0033] An online rolling optimization mechanism is adopted, which updates the prediction model weights at fixed intervals based on the latest meteorological data;

[0034] When the prediction error exceeds 10%, it automatically switches to the historical similar day pattern matching strategy;

[0035] To address sandstorm weather, a photovoltaic panel cleanliness degradation factor is introduced to correct solar energy forecast values.

[0036] Furthermore, the calculation process of the dynamic decision matrix includes:

[0037] Construct a four-dimensional state space: {energy availability, energy storage SOC, load demand, electricity price period};

[0038] The model is selected by training a strategy using the Q-learning algorithm.

[0039] Furthermore, the power allocation ratio adjustment in step 5 includes:

[0040] Generate a multi-energy coordinated control instruction set, including:

[0041] PWM duty cycle adjustment command for wind power channel;

[0042] MPPT reference voltage offset command for solar channel;

[0043] Current limit instructions for energy storage charging and discharging circuits;

[0044] By using closed-loop feedback control, the deviation between the actual power allocation result and the strategy command is kept to be ≤5%.

[0045] Furthermore, it also includes a strategy backtracking optimization mechanism:

[0046] Store historical strategy decision trees and corresponding energy utilization efficiency data;

[0047] When the same combination of environmental states is detected to occur repeatedly, the historical best strategy is used to replace the new strategy calculation.

[0048] Monthly pattern recognition is performed on strategy conflict scenarios to generate a strategy selection rule knowledge base to accelerate initial decision-making.

[0049] The substantial effects of this invention:

[0050] 1. In this invention, by constructing a closed-loop management framework of data acquisition, prediction, decision-making, execution, and re-triggering, the global optimization problem of multi-energy coordination is fundamentally solved. The adaptive strategy switching based on wind and solar power generation prediction values ​​and real-time load demand greatly improves power supply stability. The dynamic charging ratio allocation and hybrid power supply priority calling mechanism reduce the number of mode switching times and significantly reduce power device losses. Furthermore, the strategy recalculation function during load / energy changes ensures improved system response time.

[0051] 2. In this invention, a multi-objective optimization function based on a dynamic decision matrix is ​​used to achieve real-time balance between electricity purchase cost, renewable energy utilization rate, and energy storage lifespan, avoiding system imbalance caused by single-objective optimization. Combined with SOC hierarchical threshold management, it effectively prevents the risk of battery overcharging / over-discharging. The power balancing algorithm actively suppresses current surges caused by differences in wind and solar power generation, and improves energy capture efficiency with MPPT charging. The time-of-use pricing linkage mechanism prioritizes the consumption of low-priced grid electricity during off-peak periods and releases energy storage power during peak periods to reduce energy costs. The fault tolerance function monitors the electrical parameters of each module in real time, performs millisecond-level isolation and strategy reconstruction for faulty units, and ensures continuous power supply capability. The four-dimensional state-space decision model constructed by the Q-learning algorithm continuously learns from historical scenarios to improve the accuracy of scheduling strategies.

[0052] 3. In this invention, the environmental prediction model introduces a rolling optimization mechanism, dynamically correcting prediction weights based on meteorological data, significantly reducing the prediction deviation of power generation under extreme weather conditions; the dust attenuation factor specifically improves the reliability of photovoltaic output prediction; the multi-energy collaborative instruction set accurately generates PWM duty cycle, MPPT voltage offset, and charging / discharging current limit control signals, ensuring high consistency between power allocation and strategy instructions through closed-loop feedback; the health assessment model, based on fault frequency, voltage fluctuation, and aging trend data, identifies high-risk modules in advance and reduces their scheduling priority to prevent fault propagation; the strategy backtracking mechanism accelerates the matching of optimal strategies under repeated environmental conditions by accumulating a knowledge base of historical decision-making scenarios, reducing the consumption of real-time computing resources. Attached Figure Description

[0053] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art 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.

[0054] Figure 1 This is a schematic diagram of the method flow in Example 1. Detailed Implementation

[0055] To facilitate understanding of the present invention, a more detailed description is provided below with reference to the accompanying drawings and specific embodiments. It should be noted that when an element is described as being "fixed to" another element, it can be directly on the other element, or one or more intermediate elements may exist between them. When an element is described as being "connected to" another element, it can be directly connected to the other element, or one or more intermediate elements may exist between them. The terms "vertical," "horizontal," "left," "right," and similar expressions used in this specification are for illustrative purposes only.

[0056] Unless otherwise defined, all technical and scientific terms used in this specification have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. The term "and / or" as used in this specification includes any and all combinations of one or more of the associated listed items.

[0057] Example 1:

[0058] like Figure 1 As shown, a multi-mode adaptive power management method includes the following steps:

[0059] Step 1: Collect multi-source power input data in real time, including at least two of the following: wind power generation, solar power generation, grid power supply status, and standby generator status, and simultaneously acquire the energy storage unit's state of charge and load demand power.

[0060] Step 2: Based on the environmental parameter prediction model, combined with meteorological data and historical power generation curves, generate predicted values ​​for wind and solar power generation within a preset future time period;

[0061] Step 3: Construct a dynamic decision matrix. Input parameters include: current energy input priority, energy storage SOC safety threshold, load power fluctuation range, time-of-use electricity price data, and power generation forecast.

[0062] Step 4: Output the optimal energy dispatch strategy based on the dynamic decision matrix. The strategy includes:

[0063] When the total output power of renewable energy exceeds the load demand, the energy storage charging mode is executed, and the charging ratio of wind power and solar power to energy storage is dynamically allocated.

[0064] When the total output power of renewable energy is less than the load demand and the energy storage SOC is greater than the lower threshold, the hybrid power supply mode is activated, and the energy storage discharge, grid power purchase or backup generator power supply is called in the preset priority order.

[0065] When the total output power of renewable energy is less than the load demand and the energy storage SOC is less than or equal to the lower limit threshold, the system will be forced to switch to the grid / backup power dominant mode.

[0066] Step 5: Execute the energy scheduling strategy through the power allocation unit, adjust the power allocation ratio of each energy channel in real time, and trigger strategy recalculation when the load changes suddenly or the energy fluctuation exceeds the threshold.

[0067] As one implementation method, the dynamic decision matrix in step 3 is constructed in the following way:

[0068] Energy priority coefficients are defined as follows: Renewable energy > Energy storage > Power grid > Backup generators;

[0069] Set SOC hierarchical management thresholds: including emergency protection lower limit, high-efficiency charging range, and overcharge protection upper limit;

[0070] A multi-objective optimization function is established, with the optimization objectives being: minimizing grid purchase costs, maximizing renewable energy utilization, and extending energy storage lifespan. The priority of the objectives is dynamically adjusted through weighting factors.

[0071] As one implementation method, the energy storage charging mode in step 4 specifically includes:

[0072] If the instantaneous power generation difference between wind and solar energy is greater than or equal to a set threshold, the power balancing algorithm will be activated to limit the charging current of highly volatile energy sources.

[0073] If the energy storage SOC is in the high-efficiency charging range, enable maximum power point tracking charging.

[0074] If the energy storage SOC is close to SOC_H, switch to trickle charging and start the excess energy discharge circuit.

[0075] As one implementation method, the hybrid power supply mode in step 4 further includes:

[0076] During off-peak hours of time-of-use electricity pricing, priority is given to grid power supply, while during peak hours, priority is given to energy storage discharge.

[0077] When the standby generator is called, the generator output power is automatically adjusted to match the load power curve, so that it operates in the range of optimal fuel efficiency.

[0078] As one implementation method, a fault-tolerant mechanism is also included:

[0079] Real-time monitoring of abnormal voltage / current fluctuations in each energy module;

[0080] When a single energy module fails, it is automatically removed from the power supply matrix, and the energy allocation strategy is recalculated.

[0081] If the energy storage unit fails, it will switch to a bypass mode where renewable energy directly supplies the load and the grid / generator provides supplementary power.

[0082] As one implementation method, the environmental parameter prediction model in step 2 is optimized in the following way:

[0083] An online rolling optimization mechanism is adopted, which updates the prediction model weights at fixed intervals based on the latest meteorological data;

[0084] When the prediction error exceeds 10%, it automatically switches to the historical similar day pattern matching strategy;

[0085] To address sandstorm weather, a photovoltaic panel cleanliness degradation factor is introduced to correct solar energy forecast values.

[0086] As one implementation method, the calculation process of the dynamic decision matrix includes:

[0087] Construct a four-dimensional state space: {energy availability, energy storage SOC, load demand, electricity price period};

[0088] The model is selected by training a strategy using the Q-learning algorithm.

[0089] As one implementation method, the power allocation ratio adjustment in step 5 includes:

[0090] Generate a multi-energy coordinated control instruction set, including:

[0091] PWM duty cycle adjustment command for wind power channel;

[0092] MPPT reference voltage offset command for solar channel;

[0093] Current limit instructions for energy storage charging and discharging circuits;

[0094] By using closed-loop feedback control, the deviation between the actual power allocation result and the strategy command is kept to be ≤5%.

[0095] As one implementation method, a strategy backtracking optimization mechanism is also included:

[0096] Store historical strategy decision trees and corresponding energy utilization efficiency data;

[0097] When the same combination of environmental states is detected to occur repeatedly, the historical best strategy is used to replace the new strategy calculation.

[0098] Monthly pattern recognition is performed on strategy conflict scenarios to generate a strategy selection rule knowledge base to accelerate initial decision-making.

[0099] It should be noted that while the preferred embodiments of the present invention are provided in the specification and accompanying drawings, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. These embodiments are not intended to impose additional limitations on the content of the present invention; their purpose is to provide a more thorough and comprehensive understanding of the disclosure of the present invention. Furthermore, the above-described technical features can be combined with each other to form various embodiments not listed above, all of which are considered to be within the scope of the present invention specification. Moreover, those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. A multi-mode adaptive power management method, characterized in that, Includes the following steps: Step 1: Collect multi-source power input data in real time, including at least two of the following: wind power generation, solar power generation, grid power supply status, and standby generator status, and simultaneously acquire the energy storage unit's state of charge and load demand power. Step 2: Based on the environmental parameter prediction model, combined with meteorological data and historical power generation curves, generate predicted values ​​for wind and solar power generation within a preset future time period; Step 3: Construct a dynamic decision matrix. Input parameters include: current energy input priority, energy storage SOC safety threshold, load power fluctuation range, time-of-use electricity price data, and the predicted power generation value. Step 4: Output the optimal energy scheduling strategy based on the dynamic decision matrix, the strategy including: When the total output power of renewable energy exceeds the load demand, the energy storage charging mode is executed, and the charging ratio of wind power and solar power to energy storage is dynamically allocated. When the total output power of renewable energy is less than the load demand and the energy storage SOC is greater than the lower threshold, the hybrid power supply mode is activated, and the energy storage discharge, grid power purchase or backup generator power supply is called in the preset priority order. When the total output power of renewable energy is less than the load demand and the energy storage SOC is less than or equal to the lower limit threshold, the system will be forced to switch to the grid / backup power dominant mode. Step 5: Execute the energy scheduling strategy through the power allocation unit, adjust the power allocation ratio of each energy channel in real time, and trigger strategy recalculation when the load changes suddenly or the energy fluctuation exceeds the threshold.

2. The multi-mode adaptive power management method according to claim 1, characterized in that, The dynamic decision matrix in step 3 is constructed in the following way: Energy priority coefficients are defined as follows: Renewable energy > Energy storage > Power grid > Backup generators; Set SOC hierarchical management thresholds: including emergency protection lower limit, high-efficiency charging range, and overcharge protection upper limit; A multi-objective optimization function is established, with the optimization objectives being: minimizing grid purchase costs, maximizing renewable energy utilization, and extending energy storage lifespan. The priority of the objectives is dynamically adjusted through weighting factors.

3. The multi-mode adaptive power management method according to claim 1, characterized in that, The energy storage charging mode in step 4 specifically includes: If the instantaneous power generation difference between wind and solar energy is greater than or equal to a set threshold, the power balancing algorithm will be activated to limit the charging current of highly volatile energy sources. If the energy storage SOC is in the high-efficiency charging range, enable maximum power point tracking charging. If the energy storage SOC is close to SOC_H, switch to trickle charging and start the excess energy discharge circuit.

4. The multi-mode adaptive power management method according to claim 1, characterized in that, The hybrid power supply mode in step 4 further includes: During off-peak hours of time-of-use electricity pricing, priority is given to grid power supply, while during peak hours, priority is given to energy storage discharge. When the standby generator is called, the generator output power is automatically adjusted to match the load power curve, so that it operates in the range of optimal fuel efficiency.

5. The multi-mode adaptive power management method according to claim 1, characterized in that, It also includes fault tolerance mechanisms: Real-time monitoring of abnormal voltage / current fluctuations in each energy module; When a single energy module fails, it is automatically removed from the power supply matrix, and the energy allocation strategy is recalculated. If the energy storage unit fails, it will switch to a bypass mode where renewable energy directly supplies the load and the grid / generator provides supplementary power.

6. The multi-mode adaptive power management method according to claim 1, characterized in that, The environmental parameter prediction model in step 2 is optimized in the following ways: An online rolling optimization mechanism is adopted, which updates the prediction model weights at fixed intervals based on the latest meteorological data; When the prediction error exceeds 10%, it automatically switches to the historical similar day pattern matching strategy; To address sandstorm weather, a photovoltaic panel cleanliness degradation factor is introduced to correct solar energy forecast values.

7. The multi-mode adaptive power management method according to claim 1, characterized in that, The calculation process of the dynamic decision matrix includes: Construct a four-dimensional state space: {energy availability, energy storage SOC, load demand, electricity price period}; The model is selected by training a strategy using the Q-learning algorithm.

8. The multi-mode adaptive power management method according to claim 1, characterized in that, The power allocation ratio adjustment in step 5 includes: Generate a multi-energy coordinated control instruction set, including: PWM duty cycle adjustment command for wind power channel; MPPT reference voltage offset command for solar channel; Current limit instructions for energy storage charging and discharging circuits; By using closed-loop feedback control, the deviation between the actual power allocation result and the strategy command is kept to be ≤5%.

9. The multi-mode adaptive power management method according to claim 1, characterized in that, It also includes a strategy backtracking optimization mechanism: Store historical strategy decision trees and corresponding energy utilization efficiency data; When the same combination of environmental states is detected to occur repeatedly, the historical best strategy is used to replace the new strategy calculation. Monthly pattern recognition is performed on strategy conflict scenarios to generate a strategy selection rule knowledge base to accelerate initial decision-making.

Images