Energy scheduling optimization method for optical storage and charging system

Abstract

The invention relates to the technical field of smart energy management and power system automation control, in particular to an optical storage and charging system energy scheduling optimization method, which comprises the following steps: step 1, acquiring real-time state data of a heterogeneous energy storage medium in an optical storage and charging system and load demand data of a heavy truck charging station to construct a load demand vector at the current moment; the heterogeneous energy storage medium comprises a power type energy storage unit and an energy type energy storage unit; 2, presetting a power response threshold value, and constructing an initial energy distribution instruction sequence; step 3, collecting historical cycle times, charge and discharge depths and current health states of the power type energy storage unit and the energy type energy storage unit, and correcting the initial energy distribution instruction sequence to obtain a final energy scheduling strategy so as to output a control instruction to the edge computing node and the cloud platform; the overall response speed and the voltage stability of the system are improved.

Description

Technical Field

[0001] This invention relates to the field of smart energy management and power system automation control technology, specifically to an energy scheduling optimization method for a photovoltaic-storage-charging system. Background Technology

[0002] In the operation scenario of the photovoltaic-storage-charging system in the charging and swapping station of heavy trucks, the load end exhibits significant high-frequency pulse-type impact characteristics. The system is usually configured with power-type and energy-type heterogeneous energy storage media to maintain power balance. Existing energy dispatch schemes generally adopt a single time-scale model or simple rule-based logic, focusing on macroscopic energy supply and demand matching. This approach often ignores the significant differences in the physicochemical properties of heterogeneous energy storage units, making it difficult to effectively distinguish between microsecond-level transient power demand and hourly-level steady-state energy demand. Due to the failure to fully consider equipment response time constants, mechanical fatigue damage, and multi-dimensional economic value, existing methods are prone to causing time mismatch between high-frequency pulse loads and the response characteristics of hybrid energy storage systems when facing severe load fluctuations such as heavy truck startup. This results in energy-type units being forced to respond to high-frequency pulses or power-type units bearing excessive energy consumption, which not only reduces the system's ability to absorb impact loads but also exacerbates the lifespan loss of expensive energy storage equipment and reduces operational efficiency. Therefore, how to achieve refined spatiotemporal decoupling and dynamic allocation of heterogeneous energy storage resources, while ensuring the system's rapid response to high-frequency pulse loads, effectively extending equipment lifespan and improving overall economic benefits, has become an urgent technical problem to be solved. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention provides an energy scheduling optimization method for a photovoltaic-storage-charging system. Specifically, the technical solution of this invention includes: Step 1: Obtain real-time status data of heterogeneous energy storage media in the photovoltaic-energy storage-charging system and load demand data of heavy-duty truck charging and swapping stations, establish a multi-time-scale load model, identify high-frequency pulse characteristics in the load demand data, and construct the load demand vector at the current moment; wherein, heterogeneous energy storage media include power-type energy storage units and energy-type energy storage units. Step 2: Preset power response threshold, compare the load demand vector with the power response threshold, calculate the response lag parameters of power-type energy storage units and energy-type energy storage units at the current moment, obtain the heterogeneous power response matching degree, and construct the initial energy allocation command sequence. Step 3: Collect historical cycle counts, charge / discharge depths, and current health status of power-type and energy-type energy storage units to establish a battery life loss model; combine the initial energy allocation command sequence, construct an equivalent cycle life loss mitigation factor using the rainflow counting method, and obtain a multi-timescale collaborative gain index based on operational efficiency calculations. Integrate the heterogeneous power response matching degree, the equivalent cycle life loss mitigation factor, and the multi-timescale collaborative gain index to perform spatiotemporal decoupling and allocation correction on the load demand vector, obtain the corrected target power allocation vector, and correct the initial energy allocation command sequence to obtain the final energy scheduling strategy, so as to output control commands to edge computing nodes and cloud platforms.

[0004] Preferably, step one includes: S11. Through a smart gateway deployed at the edge of the photovoltaic energy storage and charging system, high-frequency data on the state of charge, terminal voltage, and temperature of power-type energy storage units and energy-type energy storage units are collected. S12. Obtain time-of-use electricity price data, spot transaction price data, and real-time power data of photovoltaic power generation system from the grid side; S13. Based on the historical load curves of heavy-duty truck charging and swapping stations, extract high-frequency pulse load characteristics and steady-state basic load characteristics, map the high-frequency pulse load characteristics to millisecond-level power demand, and map the steady-state basic load characteristics to hourly-level energy demand, and combine them to construct a load demand vector.

[0005] Preferably, step S13 includes: S131. Collect the output power waveform of the heavy-duty truck fast charging pile, identify the time period when the power ramp rate exceeds the preset ramp threshold, and mark it as the pulse impact period. S132. Collect the battery swapping frequency and battery pack quantity of heavy-duty truck battery swapping stations to predict the energy throughput demand within a preset time window in the future. S133. Calculate the system net load gap by combining the real-time power data of the photovoltaic power generation system with the real-time allowable interactive power on the grid side. S134. The power amplitude during the pulse impact period, the energy throughput demand within the future preset time window, and the system net load gap are time-aligned to generate a load demand vector.

[0006] Preferably, step two includes: S21. Extract the first response time constant of the power-type energy storage unit and the second response time constant of the energy-type energy storage unit from the physical characteristic parameters of the heterogeneous energy storage medium. S22. Preset power response threshold, which characterizes the minimum power support rate required for the system to maintain voltage stability; S23. Compare the pulse impulse component in the load demand vector with the power response threshold, calculate the deviation between the instantaneous output capability of the power storage unit and the load demand, and obtain the heterogeneous power response matching degree. S24. Based on the heterogeneous power response matching degree, the high-frequency pulse load is initially allocated to the power-type energy storage unit, and the steady-state base load is allocated to the energy-type energy storage unit, generating an initial energy allocation command sequence.

[0007] Preferably, step three includes: S31. Obtain the capacity decay curves of power-type energy storage units and energy-type energy storage units at different discharge rates, and establish a battery life loss model. S32. Input the initial energy allocation command sequence into the battery life loss model to simulate the internal temperature rise and mechanical stress changes of the battery after the allocation command is executed; S33. Use the rainflow counting method to statistically analyze the charging and discharging cycle waveforms during the simulation process, calculate the cumulative fatigue damage of each energy storage unit, and construct an equivalent cycle life loss mitigation factor. S34. Combining the time-of-use electricity price data and carbon trading price data from the grid side, calculate the economic benefits and carbon emission reduction values ​​after executing the initial energy allocation instruction sequence, and construct a multi-timescale synergistic gain index by weighted summation.

[0008] Preferably, step three also includes: S35. Extract the equivalent cycle lifetime loss mitigation factor and the multi-timescale synergistic gain index, and perform weighted correction on the power allocation ratio in the initial energy allocation command sequence. S36. When the equivalent cycle life loss mitigation factor exceeds the preset life loss warning line, reduce the proportion of power-type energy storage units and increase the real-time supplementation proportion on the grid side to obtain the corrected target power allocation vector. S37. Based on the target power allocation vector, generate the final energy dispatch strategy, which includes millisecond-level power control commands for power-type energy storage units and hourly-level energy dispatch commands for energy-type energy storage units.

[0009] Preferably, step three also includes: S38. Preset a first preset matching value and a second preset matching value, wherein the first preset matching value is greater than the second preset matching value; evaluate the heterogeneous power response matching degree to obtain the response matching evaluation result, including: If the heterogeneous power response matching degree is greater than or equal to the first preset matching value, it means that the system response speed meets the load pulse requirements, and the response matching evaluation result is marked as excellent. If the heterogeneous power response matching degree is less than the first preset matching value and greater than or equal to the second preset matching value, it indicates that the system response has a delay but has not caused a crash, and the response matching evaluation result is marked as a critical label; If the heterogeneous power response matching degree is less than the second preset matching value, it means that the system response cannot follow the load change, and the response matching evaluation result is marked as a mismatch label.

[0010] Preferably, S38 further includes: S381. For the time period marked as mismatch in the response matching evaluation results, generate the first emergency adjustment strategy, including: forcibly disconnecting some non-critical loads and activating the overload protection mode of the power storage unit. S382. For the time period marked as critical by the response matching evaluation result, generate a second smoothing adjustment strategy, including: calling the cloud platform to predict the load fluctuation at the next moment, and adjusting the basic output power of the energy storage unit in advance to reserve the adjustment margin of the power storage unit.

[0011] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention establishes a multi-timescale load model to accurately identify and separate the high-frequency pulse characteristics and steady-state basic characteristics of heavy-duty truck charging, mapping millisecond-level power demand and hourly-level energy demand to power-type and energy-type energy storage units respectively. This spatiotemporal decoupling mechanism effectively solves the time mismatch problem between impact load and heterogeneous response characteristics, avoids inefficient operating conditions caused by a single control strategy, and significantly improves the system's dynamic absorption capacity for pulse loads. 2. Based on the difference in physical response time constants of heterogeneous energy storage media, this invention constructs a matching degree calculation model based on instantaneous response deviation; by quantifying the ability of energy storage units to follow the current load, it realizes precise task allocation, prevents the low efficiency of energy-type batteries due to forced response to high-frequency pulses, ensures that each energy storage unit operates under optimal physical conditions, and improves the overall response speed and voltage stability of the system. 3. This invention introduces a thermo-coupling model and rainflow counting method to transform the charging and discharging waveform into a quantifiable mechanical fatigue damage index and construct a life loss mitigation factor; it integrates equipment health constraints into the allocation strategy, automatically reducing the proportion of power-type units and introducing grid compensation when losses are too high; this not only prevents premature equipment failure due to excessive pursuit of response speed, but also effectively balances instantaneous performance and long-term lifespan, extending the service life of expensive energy storage assets; 4. This invention comprehensively considers time-of-use electricity pricing, carbon trading prices, and operational efficiency to construct a multi-time-scale synergistic gain index and establish a graded response assessment and emergency adjustment mechanism. By weighted correction of initial instructions and proactive intervention under mismatch conditions, a dynamic balance between economic benefits, carbon emission reduction, and operational safety is achieved. While ensuring the power supply reliability of heavy-duty truck charging and battery swapping stations, the comprehensive operational value of the photovoltaic-storage-charging system is maximized. Attached Figure Description

[0012] The present invention will be further explained below with reference to the accompanying drawings and embodiments: Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0013] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0014] Example 1: Please see Figure 1 A method for optimizing energy scheduling in a photovoltaic energy storage and charging system, comprising the following steps: Step 1: Obtain real-time status data of heterogeneous energy storage media in the photovoltaic-energy storage-charging system and load demand data of heavy-duty truck charging and swapping stations, establish a multi-time-scale load model, identify high-frequency pulse characteristics in the load demand data, and construct the load demand vector at the current moment; wherein, heterogeneous energy storage media include power-type energy storage units and energy-type energy storage units. Step 2: Preset power response threshold, compare the load demand vector with the power response threshold, calculate the response lag parameters of power-type energy storage units and energy-type energy storage units at the current moment, obtain the heterogeneous power response matching degree, and construct the initial energy allocation command sequence. Step 3: Collect historical cycle counts, charge / discharge depths, and current health status of power-type and energy-type energy storage units to establish a battery life loss model; combine the initial energy allocation command sequence, construct an equivalent cycle life loss mitigation factor using the rainflow counting method, and obtain a multi-timescale collaborative gain index based on operational efficiency calculations. Integrate the heterogeneous power response matching degree, the equivalent cycle life loss mitigation factor, and the multi-timescale collaborative gain index to perform spatiotemporal decoupling and allocation correction on the load demand vector, obtain the corrected target power allocation vector, and correct the initial energy allocation command sequence to obtain the final energy scheduling strategy, so as to output control commands to edge computing nodes and cloud platforms.

[0015] This embodiment details the execution of the energy scheduling optimization method for the photovoltaic-storage-charging system described above, aiming to solve the timing mismatch problem between the high-frequency pulsed impact load unique to heavy-duty truck charging and battery swapping stations and the heterogeneous response characteristics of the hybrid energy storage system. This method employs a dynamic scheduling scheme based on spatiotemporal decoupling, and the specific execution logic is as follows: The system establishes a multi-timescale load model, a mathematical framework capable of simultaneously characterizing microsecond-level transient changes and hourly-level steady-state trends. Its core objective is to separate the random pulse behavior of heavy-duty truck charging from the background load. During this process, the system acquires real-time state data of heterogeneous energy storage media within the photovoltaic-energy storage-charging system. These heterogeneous energy storage media specifically refer to two types of energy storage units with drastically different physicochemical properties: power-type energy storage units use lithium iron phosphate batteries, possessing high power density and millisecond-level response speeds, but with relatively short cycle lives; energy-type energy storage units use vanadium redox flow batteries, possessing long cycle lives and deep charge / discharge capabilities, but limited by the pump system, with response speeds ranging from seconds to minutes. The system identifies high-frequency pulse characteristics in the load demand data, such as the power spike at the moment the heavy-duty truck starts fast charging, and constructs the current time-of-flight state data. Load demand vector ; The system performs calculations of heterogeneous power response matching and constructs initial instructions. The core of this step lies in quantifying the energy storage medium's ability to follow the current load. The system presets a power response threshold. This threshold defines the critical point of the load change rate, by... and By comparison, the response hysteresis parameters of the two types of energy storage units are calculated. To quantify this hysteresis, this embodiment employs a calculation model based on instantaneous response deviation and defines the heterogeneous power response matching degree. The calculation formula is as follows: in, To adjust the dimensionless weighting coefficient of sensitivity, As the system reference power, This refers to the maximum output power of a power-type energy storage unit at the current moment, constrained by its state of charge, temperature, and health status. The duration of the response after the pulse is generated. The first response time constant of the power-type energy storage unit. The high-frequency pulse component of the load; The system incorporates the dual constraints of battery lifespan degradation and economic benefits, establishing a battery lifespan degradation model, and combining it with the initial energy allocation command sequence. Using rainflow counting method, construct an equivalent cycle life loss mitigation factor. Simultaneously, obtain the multi-timescale synergistic gain index based on operational efficiency calculations. This index integrates economic returns across different time scales; systematic integration , and ,right Spatiotemporal decoupling is implemented: in the spatial dimension, the total power demand is decomposed into power-type and energy-type units with different physical locations; in the temporal dimension, instantaneous power support tasks are separated from long-term energy transfer tasks; through correction The target power allocation vector is obtained. It also outputs control commands to edge computing nodes and the cloud platform; This embodiment achieves refined utilization of heterogeneous energy storage resources by comprehensively considering physical response characteristics, equipment health, and economic value. In high-frequency impact load scenarios such as heavy-duty truck charging and swapping stations, this strategy effectively avoids inefficient operating conditions caused by a single control strategy, such as over-engineering or under-engineering, improves the system's ability to absorb pulse loads, and extends the service life of expensive energy storage assets.

[0016] Example 2: Step one includes: S11. Through a smart gateway deployed at the edge of the photovoltaic energy storage and charging system, high-frequency data on the state of charge, terminal voltage, and temperature of power-type energy storage units and energy-type energy storage units are collected. S12. Obtain time-of-use electricity price data, spot transaction price data, and real-time power data of photovoltaic power generation system from the grid side; S13. Based on the historical load curves of heavy-duty truck charging and swapping stations, extract high-frequency pulse load characteristics and steady-state basic load characteristics, map the high-frequency pulse load characteristics to millisecond-level power demand, and map the steady-state basic load characteristics to hourly-level energy demand, and combine them to construct a load demand vector.

[0017] This embodiment further defines the data acquisition and feature extraction process in step one; the system executes a multi-source state perception step, acquiring the state of charge of power-type and energy-type energy storage units at a high-frequency sampling rate of 100Hz through a smart gateway deployed at the edge. Terminal voltage and temperature The state of charge This refers to the ratio of the battery's remaining charge to its rated capacity, serving as a basis for determining whether the battery is overcharged or over-discharged. During this period, the system simultaneously interacts with the power grid and photovoltaic data to obtain time-of-use electricity price data from the grid side. Spot transaction price and the real-time power of the photovoltaic power generation system These data provide boundary conditions for subsequent economic assessments; The system performs a two-layer mapping of load characteristics. Based on the historical load curves of heavy-duty truck charging and battery swapping stations, it extracts two distinct characteristics: first, high-frequency pulse load characteristics, corresponding to the startup and power ramp-up phases of heavy-duty truck fast charging piles, which the system maps to millisecond-level power demand. This type of demand has low energy requirements but extremely high power response speed requirements; second, steady-state base load characteristics, corresponding to the station's lighting, air conditioning, and the slow charging process of the battery swapping station's battery packs, which the system maps to hourly-level energy demand. The system constructs a load demand vector by combining these two types of demand. ; This embodiment decomposes load characteristics into two dimensions: millisecond-level power and hourly-level energy, providing a data foundation for subsequent precise allocation of energy storage media with different characteristics. In the grid interaction scenario, this two-layer mapping ensures that the granularity of the data source meets the control accuracy requirements, enabling the system to distinguish between instantaneous impacts and continuous energy consumption, thereby optimizing the allocation of computing resources.

[0018] Example 3: Step S13 includes: S131. Collect the output power waveform of the heavy-duty truck fast charging pile, identify the time period when the power ramp rate exceeds the preset ramp threshold, and mark it as the pulse impact period. S132. Collect the battery swapping frequency and battery pack quantity of heavy-duty truck battery swapping stations to predict the energy throughput demand within a preset time window in the future. S133. Calculate the system net load gap by combining the real-time power data of the photovoltaic power generation system with the real-time allowable interactive power on the grid side. S134. The power amplitude during the pulse impact period, the energy throughput demand within the future preset time window, and the system net load gap are time-aligned to generate a load demand vector.

[0019] This embodiment refines the load demand vector construction process in step S13; the system performs pulse impact period identification, collects the output power waveform of the heavy-duty truck fast charging pile, and calculates the power ramp-up rate. The calculation formula is as follows: in, For load power, For time response Exceeding the preset climbing threshold For example, if the speed is 50 kW / s, the system marks this period as a pulse impact period; the system performs energy throughput prediction and collects the battery swapping frequency of the battery swapping station. With the number of battery packs Predicting energy throughput demand within a preset time window in the future. The system calculates the net load gap. The calculation formula is as follows: in, Real-time photovoltaic power. The system aligns the above parameters in time to generate a load demand vector for the maximum allowable interactive power of the power grid in real time. ,in This represents the power amplitude during the pulse impact period; This embodiment clarifies the specific composition of the load vector, which includes not only the amount of power shortage but also the form of power shortage. In integrated photovoltaic, energy storage, and charging stations, this vectorized expression enables the scheduling algorithm to distinguish between instantaneous surges and continuous throughput, thereby avoiding misjudgments caused by photovoltaic fluctuations and improving the system's dynamic adaptability to changes in net load.

[0020] Example 4: Step two includes: S21, extracting the first response time constant of the power-type energy storage unit and the second response time constant of the energy-type energy storage unit from the physical characteristic parameters of the heterogeneous energy storage medium; S22. Preset power response threshold, which characterizes the minimum power support rate required for the system to maintain voltage stability; S23. Compare the pulse impulse component in the load demand vector with the power response threshold, calculate the deviation between the instantaneous output capability of the power storage unit and the load demand, and obtain the heterogeneous power response matching degree. S24. Based on the heterogeneous power response matching degree, the high-frequency pulse load is initially allocated to the power-type energy storage unit, and the steady-state base load is allocated to the energy-type energy storage unit, generating an initial energy allocation command sequence.

[0021] This embodiment refines the heterogeneous power response matching degree calculation and initial allocation in step two; the system extracts the response time constant and obtains the first response time constant of the power-type energy storage unit from the physical characteristic parameters. Its time constant is on the order of 10-100 ms, which is similar to the second response time constant of an energy storage unit. Its duration is on the order of 1-10 seconds; the system includes timer reset logic; the system monitors the load change rate in real time. ,when When a pulse event is triggered, the timer is set to... Reset to And begin accumulating time; when If the timer continues to stabilize beyond the preset time, stop the timer and reset. For infinity, such that This ensures the accuracy of dynamic response calculation; based on Execute initial command generation to initially generate high-frequency pulsed load. The power-type energy storage unit will allocate the steady-state base load. Allocate to energy storage units and generate an initial energy allocation command sequence. ; This embodiment utilizes the physical time constant. As the basis for allocation, it fundamentally follows the physical characteristics of the equipment; in the actual operation of the hybrid energy storage system, this time constant-based allocation strategy prevents the energy-type batteries from becoming inefficient and losing their lifespan due to forced response to high-frequency pulses, and ensures that each energy storage unit operates under its own optimal conditions.

[0022] Example 5: Step 3 includes: S31, obtaining the capacity decay curves of power-type energy storage units and energy-type energy storage units at different discharge rates, and establishing a battery life loss model; S32. Input the initial energy allocation command sequence into the battery life loss model to simulate the internal temperature rise and mechanical stress changes of the battery after the allocation command is executed; S33. Use the rainflow counting method to statistically analyze the charging and discharging cycle waveforms during the simulation process, calculate the cumulative fatigue damage of each energy storage unit, and construct an equivalent cycle life loss mitigation factor. S34. Combining the time-of-use electricity price data and carbon trading price data from the grid side, calculate the economic benefits and carbon emission reduction values ​​after executing the initial energy allocation instruction sequence, and construct a multi-timescale synergistic gain index by weighted summation.

[0023] This embodiment refines the process of constructing multidimensional evaluation factors in step three; the system establishes a battery life loss model; in this step, in addition to obtaining the capacity decay curves of power-type and energy-type energy storage units at different discharge rates, the system also reads the battery thermal capacity from the system's preset equipment physical parameter library. Young's modulus Poisson's ratio and partial molar volume Basic parameters; the capacity decay curve is mainly used to calibrate and correct the stress-lifetime mapping relationship in the model, specifically using a multiplicative gain correction strategy: the system adjusts the current capacity retention rate... Using the formula: in, The aging hardening coefficient, for example, can be 1.5 to 2.0 for lithium iron phosphate batteries. This coefficient is derived by fitting mechanical fatigue test data of battery materials to reflect the changes in mechanical properties caused by battery aging, so as to ensure the accuracy of the model at different aging stages. For step S32, the initial energy allocation command sequence only contains power commands. To address the issue of the physical model requiring current input, the system supplements this by establishing a voltage-current conversion model: pre-setting the battery's open-circuit voltage-state of charge (OCV-SOC) mapping curve, along with the equivalent internal resistance. By solving the equations, the real-time current is calculated. And update in real time using the ampere-hour integration method. This provides accurate current input for subsequent thermal and mechanical models. The calculation formula is as follows: The system inputs the initial energy distribution command sequence into the model to simulate the changes in internal temperature rise and mechanical stress of the battery; among which, the initial boundary conditions of the simulation process... and The temperature data and state of charge data collected in real time in step S11 are used respectively; in order to accurately quantify the above-mentioned physical field changes, this embodiment discloses specific model equations: for the internal temperature rise, a lumped-parameter thermal model is used for calculation: in, The real-time current flowing through the energy storage unit, For battery heat capacity, For internal resistance, The convective heat transfer coefficient is... For heat dissipation area, The ambient temperature is used; this formula can calculate the battery core temperature in real time after the command is executed. For mechanical stress, a mechanical model based on diffusion-induced stress is used for calculation; in order to construct the input current... To map the relationship between the lithium ion concentration distribution inside the particles, this embodiment introduces the spherical coordinate form of Fick's second law as the equation of state, and its calculation formula is as follows: in, For time, The coordinates represent the radial distance within the active particles. Boundary conditions are defined as follows: in, Let be the solid-phase diffusion coefficient. Where is the particle radius, It is Faraday's constant. This represents the total effective electrochemical reaction area inside the battery. By discretizing and solving the above equations using the finite difference method, the particle surface concentration can be obtained in real time. With volume average concentration Substituting it into the stress formula: in, For diffusion-induced stress, For partial molar volume, for example, taking the graphite anode material... This is the corrected Young's modulus. Poisson's ratio, and These represent the particle surface concentration and the average concentration, respectively; this model can simulate the particle expansion and contraction stress during the charging and discharging process. The system uses rainflow counting to statistically analyze the diffusion-induced stress during the simulation process. The time-domain waveform is used here, specifically the stress waveform instead of the current waveform, to ensure logical consistency with the preceding mechanical model, identifying the number of full and half cycles and their corresponding stress amplitudes. Based on this, the system calculates the equivalent cycle life loss mitigation factor. The calculation formula is as follows: in, The lifespan depreciation weighting factor takes into account different battery lifespans. The difference in damage sensitivity between intervals is set to a range of values. In this embodiment, the default value is ; The total number of groups of stress cycles with different amplitudes identified by the rainflow counting method; For the first The actual number of stress cycles; For the corresponding stress amplitude The theoretical fatigue life count is given; to solve the logical closed-loop problem in variable calculation, this embodiment explicitly discloses... With stress amplitude The functional relationship between them, i.e., the SN fatigue curve: It should be noted that in the formula The original stress amplitude is expressed in Pascals (Pa). Used to calculate the original stress amplitude in megapascals (MPa) within the formula, term Used to perform unit conversion to megapascals within the formula to match the fatigue strength coefficient. The dimensions of; among which, For example, the fatigue strength coefficient of the material. , The fatigue ductility index, for example, 3.5. From The first extracted from the waveform The stress amplitude converted to MPa for each cycle; This factor Defined as a quantitative indicator of cumulative fatigue damage, the higher the value (closer to or exceeding 1), the more severe the battery life loss caused by the current allocation strategy, i.e., the greater the loss. It should be noted that the values ​​calculated in this step... It is the fatigue damage increment generated based on the current initial energy allocation instruction sequence, i.e., a single scheduling window, rather than the total cumulative value over the entire battery life cycle. Therefore, its order of magnitude is usually much smaller than 1. To assess whether this loss is excessive in subsequent correction steps, it needs to be compared with the loss allowance permitted within that time window, as discussed later. Comparison; System construction of multi-timescale cooperative gain exponent By combining time-of-use electricity price data from the power grid side with carbon trading price data, the execution... Subsequent economic benefits With carbon emission reduction To ensure the feasibility of this step, this embodiment discloses a specific computational mapping model, the calculation formula of which is as follows: in, The time interval for a single time step. The total number of time steps within the scheduling window. This is the index of the time step within the current scheduling window; These represent the discharge and charging power of the energy storage unit, respectively. For time-of-use electricity pricing, This is a fixed value representing the estimated savings in electricity costs within this scheduling window based on historical data. For direct photovoltaic power consumption, Carbon emission factors of grid electricity, for example ; To eliminate dimensional differences, the maximum economic return within the historical statistical period is used. With maximum carbon emission reduction The above indicators are normalized to obtain the normalized economic returns. With normalized carbon emission reduction Using formulas Perform a weighted summation; These are dimensionless weighting coefficients; This embodiment introduces a thermo-mechanical coupling model and rainflow counting method to transform complex and irregular charging and discharging waveforms into quantifiable fatigue damage indicators. In the scenario of battery life cycle management, this method makes the optimization target no longer limited to the immediate power balance, but extends to the long-term health of the equipment, achieving a balance between economic benefits and asset protection.

[0024] Example 6: Step 3 also includes: S35, extracting the equivalent cycle lifetime loss mitigation factor and the multi-timescale synergistic gain index, and weighting and correcting the power allocation ratio in the initial energy allocation command sequence; S36. When the equivalent cycle life loss mitigation factor exceeds the preset life loss warning line, reduce the proportion of power-type energy storage units and increase the real-time supplementation proportion on the grid side to obtain the corrected target power allocation vector. S37. Based on the target power allocation vector, generate the final energy dispatch strategy, which includes millisecond-level power control commands for power-type energy storage units and hourly-level energy dispatch commands for energy-type energy storage units.

[0025] This embodiment is a refinement of the allocation correction and final strategy generation in step three; the system performs weighted correction and extracts... and And simultaneously call the heterogeneous power response matching degree calculated in step two. Construct correction coefficients ,right The power allocation ratio in the middle is corrected; in order to solve That is, lifespan loss, That is, economic gains and In response to the problem of inconsistent matching dimensions, and to strictly implement the technical requirement of comprehensively correcting the three factors in claim 1, this embodiment discloses... The specific calculation formula is as follows: in, To match the weighting coefficients, for example, values ​​ranging from 0.5 to 1.0, are used to strengthen the constraint of physical response characteristics on the allocation strategy; : Non-negativity constraint operator to prevent the correction coefficient from becoming negative due to excessive instantaneous loss; k The loss penalty factor is preferably set to 1.0 to ensure that power allocation only stops when the loss reaches its limit. The set lifetime loss limit is used as a normalization benchmark, specifically referring to the maximum allowable fatigue damage increment under the current scheduling time scale; In order to unify the physical dimensions of fatigue damage, The calculation formula is revised as follows: in, The total number of cycles is designed for the entire lifecycle of the energy storage unit; here, Specifically, it refers to the total number of design cycle lives of the energy storage unit measured under standard rated operating conditions, such as 1C charge / discharge rate and 25℃ constant temperature environment, as a benchmark reference value for damage assessment; This represents the theoretically maximum number of charge-discharge cycles allowed within the current scheduling time window. The safety margin factor is used to retain a design margin in life prediction, and its value typically ranges from 0.8 to 0.95; in this embodiment, it is set to 0.9. This formula modifies the life evaluation benchmark from a time dimension to a cycle count dimension to match the cumulative fatigue damage calculated by the rainflow counting method. The physical meaning; The formula clarifies The linear correlation with the time window length prevents damage caused by changes in the time scale. Numerical distortion; This is an economic incentive factor, for example, with a value of 0.1 to 0.3, used to appropriately increase power output when economic benefits are high; Using the arctangent function and normalization coefficients Economic gain index Map to the normalized interval to eliminate the influence of dimensions; It should be noted that, in order to resolve the issues of variable dimension mismatch and time domain conflict, this embodiment explicitly defines: ψ is a scalar constant obtained through offline statistical calculation of the initial instruction sequence A0 of the current scheduling window, which is predicted based on the current time and is generated at the current moment. This sequence, such as a 15-minute sequence, remains unchanged within the real-time control period of the current window. It is over time Instantaneous variables changing in milliseconds; correction coefficient The computational logic essentially utilizes long-period statistical characteristics, namely lifespan and economy, as static gains to respond to short-period instantaneous response variables. Perform dynamic modulation; Calculated using this formula Then, execute The system activates its lifespan protection mechanism and sets a lifespan depletion warning threshold. ; in response This means that the loss factor has exceeded the safety threshold, i.e., the accumulated damage is too great. The system forcibly triggers protection logic, reducing the load on power-type energy storage units and increasing the real-time replenishment ratio on the grid side. Initial energy allocation instruction sequence The pre-allocated power of a medium-power energy storage unit at the current moment; The following power compensation formula is specifically executed to achieve a closed-loop logic: That is, the power share of power-type energy storage units that is reduced due to lifespan protection. The load is transferred to the grid side in real time, thereby ensuring the conservation of the total power of the system. Thus, the corrected target power allocation vector is obtained. ; The system is based on The final energy dispatch strategy is generated, which consists of two levels: first, millisecond-level power control commands sent to edge nodes to control LFP batteries and DC / DC converters to smooth out instantaneous fluctuations; second, hourly-level energy dispatch commands sent to the cloud platform to control the electrolyte pump speed and charge / discharge schedule of VRB batteries to achieve peak shaving and valley filling. This embodiment implements closed-loop control. When pursuing high performance leads to excessively rapid battery degradation, i.e. When the value is too high, the algorithm will automatically give way and introduce grid support. In heavy truck charging stations where extreme conditions occur frequently, this mechanism finds a dynamic balance between performance and lifespan, preventing premature equipment failure due to excessive pursuit of response speed.

[0026] Example 7: Step 3 also includes: S38, setting a first preset matching value and a second preset matching value, wherein the first preset matching value is greater than the second preset matching value; evaluating the heterogeneous power response matching degree to obtain a response matching evaluation result, including: if the heterogeneous power response matching degree is greater than or equal to the first preset matching value, it indicates that the system response speed meets the load pulse requirements, and the response matching evaluation result is marked as excellent; If the heterogeneous power response matching degree is less than the first preset matching value but greater than or equal to the second preset matching value, it indicates that the system response has a delay but has not caused a crash, and the response matching evaluation result is marked as a critical label; if the heterogeneous power response matching degree is less than the second preset matching value, it indicates that the system response cannot follow the load change, and the response matching evaluation result is marked as a mismatch label.

[0027] This embodiment refines the system response status evaluation; the system performs multi-level matching evaluation, with a preset first preset matching value. Matching the second preset value System response matching degree to heterogeneous power Conduct an assessment: in response to The system determines that it fully follows the load and marks it as excellent; in response to The system determines that there is a response delay but no voltage collapse, and marks it as a critical tag; in response to The system determined that the response was severely delayed and marked it as a mismatch. This embodiment establishes a hierarchical evaluation mechanism, enabling the system to perceive its own health and competence in real time. In the complex power grid interaction environment, this state perception capability provides a decision-making basis for triggering emergency strategies of different levels, ensuring the operational safety of the system under different operating conditions.

[0028] Example 8: S38 also includes: S381, generating a first emergency adjustment strategy for the period marked as mismatch in the response matching evaluation results, including: forcibly disconnecting some non-critical loads and activating the overload protection mode of the power storage unit. S382. For the time period marked as critical by the response matching evaluation result, generate a second smoothing adjustment strategy, including: calling the cloud platform to predict the load fluctuation at the next moment, and adjusting the basic output power of the energy storage unit in advance to reserve the adjustment margin of the power storage unit.

[0029] This embodiment is a refinement of the adjustment strategy under different evaluation results; for the time period marked as mismatch in response matching evaluation results, the system generates a first emergency adjustment strategy, which includes forcibly disconnecting some non-critical loads and activating the overload protection mode of the power storage unit, allowing it to output 1.2 times the rated power within a preset overload allowable time window to prevent system collapse. For the time periods marked as critical by the response matching evaluation results, the system generates a second smoothing adjustment strategy and calls the cloud platform to predict the next moment. To mitigate load fluctuations, adjust the base output power of the energy storage unit in advance, for example, by increasing the output of the VRB 5 seconds in advance, thereby providing more adjustment margin for the power storage unit and enabling it to better cope with upcoming pulses; This embodiment provides a fault-tolerance mechanism for the system; under extreme operating conditions, by sacrificing local experience or pre-drawing future capabilities, it ensures the safe and stable operation of the entire heavy-duty truck charging and swapping station system, effectively avoids downtime accidents caused by response mismatch, and improves the power supply reliability on the grid side.

[0030] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for optimizing energy scheduling in a photovoltaic-storage-charging system, characterized in that, The specific steps include: Step 1: Obtain real-time status data of heterogeneous energy storage media in the photovoltaic-energy storage-charging system and load demand data of heavy-duty truck charging and swapping stations, establish a multi-time-scale load model, identify high-frequency pulse characteristics in the load demand data, and construct the load demand vector at the current moment; wherein, the heterogeneous energy storage media includes power-type energy storage units and energy-type energy storage units. Step 2: Preset power response threshold, compare the load demand vector with the power response threshold, calculate the response lag parameters of power-type energy storage units and energy-type energy storage units at the current moment, obtain the heterogeneous power response matching degree, and construct the initial energy allocation command sequence. Step 3: Collect historical cycle counts, charge / discharge depths, and current health status of power-type and energy-type energy storage units to establish a battery life loss model; combine the initial energy allocation command sequence, construct an equivalent cycle life loss mitigation factor using the rainflow counting method, and obtain a multi-timescale collaborative gain index based on operational efficiency calculations. Integrate the heterogeneous power response matching degree, the equivalent cycle life loss mitigation factor, and the multi-timescale collaborative gain index to perform spatiotemporal decoupling and allocation correction on the load demand vector, obtain the corrected target power allocation vector, and correct the initial energy allocation command sequence to obtain the final energy scheduling strategy, so as to output control commands to edge computing nodes and cloud platforms.

2. The energy scheduling optimization method for a photovoltaic energy storage and charging system according to claim 1, characterized in that, Step one includes: S11. Through a smart gateway deployed at the edge of the photovoltaic energy storage and charging system, high-frequency data on the state of charge, terminal voltage, and temperature of power-type energy storage units and energy-type energy storage units are collected. S12. Obtain time-of-use electricity price data, spot transaction price data, and real-time power data of photovoltaic power generation system from the grid side; S13. Based on the historical load curves of heavy-duty truck charging and swapping stations, extract high-frequency pulse load characteristics and steady-state basic load characteristics, map the high-frequency pulse load characteristics to millisecond-level power demand, and map the steady-state basic load characteristics to hourly-level energy demand, and combine them to construct a load demand vector.

3. The energy scheduling optimization method for a photovoltaic energy storage and charging system according to claim 2, characterized in that, Step S13 includes: S131. Collect the output power waveform of the heavy-duty truck fast charging pile, identify the time period when the power ramp rate exceeds the preset ramp threshold, and mark it as the pulse impact period. S132. Collect the battery swapping frequency and battery pack quantity of heavy-duty truck battery swapping stations to predict the energy throughput demand within a preset time window in the future. S133. Calculate the system net load gap by combining the real-time power data of the photovoltaic power generation system with the real-time allowable interactive power on the grid side. S134. The power amplitude during the pulse impact period, the energy throughput demand within the future preset time window, and the system net load gap are time-aligned to generate a load demand vector.

4. The energy scheduling optimization method for a photovoltaic energy storage and charging system according to claim 3, characterized in that, Step two includes: S21. Extract the first response time constant of the power-type energy storage unit and the second response time constant of the energy-type energy storage unit from the physical characteristic parameters of the heterogeneous energy storage medium. S22. Preset power response threshold, which characterizes the minimum power support rate required for the system to maintain voltage stability; S23. Compare the pulse impulse component in the load demand vector with the power response threshold, calculate the deviation between the instantaneous output capability of the power storage unit and the load demand, and obtain the heterogeneous power response matching degree. S24. Based on the heterogeneous power response matching degree, the high-frequency pulse load is initially allocated to the power-type energy storage unit, and the steady-state base load is allocated to the energy-type energy storage unit, generating an initial energy allocation command sequence.

5. The energy scheduling optimization method for a photovoltaic energy storage and charging system according to claim 4, characterized in that, Step three includes: S31. Obtain the capacity decay curves of power-type energy storage units and energy-type energy storage units at different discharge rates, and establish a battery life loss model. S32. Input the initial energy allocation command sequence into the battery life loss model to simulate the internal temperature rise and mechanical stress changes of the battery after the allocation command is executed; S33. Use the rainflow counting method to statistically analyze the charging and discharging cycle waveforms during the simulation process, calculate the cumulative fatigue damage of each energy storage unit, and construct an equivalent cycle life loss mitigation factor. S34. Combining the time-of-use electricity price data and carbon trading price data from the grid side, calculate the economic benefits and carbon emission reduction values ​​after executing the initial energy allocation instruction sequence, and construct a multi-timescale synergistic gain index by weighted summation.

6. The energy scheduling optimization method for a photovoltaic energy storage and charging system according to claim 5, characterized in that, Step three also includes: S35. Extract the equivalent cycle lifetime loss mitigation factor and the multi-timescale synergistic gain index, and perform weighted correction on the power allocation ratio in the initial energy allocation command sequence. S36. When the equivalent cycle life loss mitigation factor exceeds the preset life loss warning line, reduce the proportion of power-type energy storage units and increase the real-time supplementation proportion on the grid side to obtain the corrected target power allocation vector. S37. Based on the target power allocation vector, generate the final energy dispatch strategy, which includes millisecond-level power control commands for power-type energy storage units and hourly-level energy dispatch commands for energy-type energy storage units.

7. The energy scheduling optimization method for a photovoltaic energy storage and charging system according to claim 6, characterized in that, Step three also includes: S38. Preset a first preset matching value and a second preset matching value, wherein the first preset matching value is greater than the second preset matching value; evaluate the heterogeneous power response matching degree to obtain the response matching evaluation result, including: If the heterogeneous power response matching degree is greater than or equal to the first preset matching value, it means that the system response speed meets the load pulse requirements, and the response matching evaluation result is marked as excellent. If the heterogeneous power response matching degree is less than the first preset matching value and greater than or equal to the second preset matching value, it indicates that the system response has a delay but has not caused a crash, and the response matching evaluation result is marked as a critical label. If the heterogeneous power response matching degree is less than the second preset matching value, it means that the system response cannot follow the load change, and the response matching evaluation result is marked as a mismatch label.

8. The energy scheduling optimization method for a photovoltaic energy storage and charging system according to claim 7, characterized in that, S38 also includes: S381. For the time period marked as mismatch in the response matching evaluation results, generate the first emergency adjustment strategy, including: forcibly disconnecting some non-critical loads and activating the overload protection mode of the power storage unit. S382. For the time period marked as critical by the response matching evaluation result, generate a second smoothing adjustment strategy, including: calling the cloud platform to predict the load fluctuation at the next moment, and adjusting the basic output power of the energy storage unit in advance to reserve the adjustment margin of the power storage unit.

Images