Scenic spot DC building micro-grid unit control system based on multi-objective optimization

Abstract

The invention discloses a scenic spot direct-current building micro-grid unit control system based on multi-objective optimization, and particularly relates to the technical field of scenic spot direct-current building micro-grid unit control, which comprises a state sensing module, a collaborative decision module, a control instruction generation module and an execution and feedback module. The state sensing module collects photovoltaic data, fan output data, energy storage SOC data, bus voltage data, load data and electricity price data. The collaborative decision-making module quickly responds to a system state through rule reasoning and outputs a preliminary equipment action, further comprehensively optimizes system economic operation, renewable energy consumption rate and energy storage life targets by utilizing a multi-target optimization model, and selects a final strategy from a non-dominated solution set based on an entropy weight method; the control instruction generation module analyzes the time sequence instruction into a time sequence instruction; and the execution and feedback module executes the instruction and updates the system state in a closed-loop manner. The system effectively improves the economy, the power supply reliability and the energy utilization efficiency of the scenic spot DC micro-grid, and prolongs the service life of the energy storage equipment.

Description

Technical Field

[0001] This invention relates to the field of DC building microgrid unit control technology in scenic areas, and more specifically, to a DC building microgrid unit control system for scenic areas based on multi-objective optimization. Background Technology

[0002] In recent years, with the rapid development of renewable energy technologies and the continuous growth of energy demand in scenic areas, DC building microgrids have been widely used in scenic area power supply systems due to their advantages such as easy access to distributed power sources and reduced AC / DC conversion losses. Currently, most common DC microgrid control strategies adopt rule-based single-objective optimization methods. However, these methods still have some drawbacks in practical applications. For example, although such methods are simple to implement, they are difficult to balance multiple objectives such as system operation economy, energy utilization efficiency, and equipment lifespan. This often leads to problems such as sacrificing power supply reliability in pursuit of low cost or accelerating the aging of energy storage units in order to improve absorption rate. In addition, existing systems have lag in dynamic response and mostly rely on periodic optimization calculations, making it difficult to respond to rapid fluctuations in scenic area load and renewable energy output in a timely manner, which can easily cause bus voltage over-limit or power outages for critical loads.

[0003] On the other hand, existing control strategies generally lack adaptability. Their rule thresholds and optimization model parameters mostly rely on manual settings and remain fixed, making it difficult to adjust them online according to the actual operating status of the system. Over long-term operation, this can easily lead to a decline in control performance, making it unable to adapt to the differentiated operational needs of scenic spots under different seasons and visitor flow scenarios. Summary of the Invention

[0004] To overcome the aforementioned deficiencies in the prior art, embodiments of the present invention provide a scenic area DC building microgrid unit control system based on multi-objective optimization. The system addresses the significant shortcomings in multi-objective collaborative optimization, rapid dynamic response, and online adaptive adjustment mentioned in the background art through the following solutions.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a multi-objective optimization-based DC building microgrid unit control system for scenic areas, comprising: Status awareness module: used to acquire real-time operating parameters of distributed power sources, energy storage units and loads in DC microgrid; Collaborative decision-making module: includes: The optimization model unit is pre-set with a multi-objective optimization function that includes objectives of economic operation, energy efficiency and power supply reliability, as well as the multi-objective optimization function and its corresponding constraints. The rule reasoning unit is pre-loaded with a fast-response rule base based on system state identification; The collaborative decision-making module is configured to: identify the current operating state by the rule reasoning unit and output a preliminary combination of equipment actions; then, use the preliminary combination of equipment actions as the initial solution and input iteratively to the optimization model unit to solve for the non-dominated solution set that satisfies the multi-objective optimization function and constraints; finally, select a final control strategy from the non-dominated solution set based on the preset decision criteria. Control command generation module: used to parse the final control strategy into timed device control commands, so as to control the scenic area DC building microgrid unit to perform tasks based on the device control commands; Execution and feedback module: used to execute the control commands and collect system response data to update the input of the state perception module, forming a closed-loop control.

[0006] The technical effects and advantages of this invention are as follows: 1. This invention effectively solves the problem that existing single-objective control strategies cannot simultaneously take into account economy, greenness and equipment durability by constructing a multi-objective optimization function that includes economic operation, renewable energy absorption rate and energy storage life, and adopts a decision-making mechanism that combines rule reasoning and optimization model, thus significantly improving the overall energy efficiency of the system. 2. This invention achieves millisecond-level preliminary decision-making through a joint state rule base based on system net power and energy storage SOC, and performs fine-tuning of the preliminary strategy by combining multi-objective optimization, thereby realizing rapid response to system state fluctuations and smooth optimization of energy storage power and load switching timing, effectively suppressing bus voltage fluctuations and ensuring the reliability of power supply to critical loads. 3. This invention collects response data such as critical load interruption duration, energy storage cycle depth, and power curtailment in real time through the execution and feedback module, and corrects the rule thresholds and optimization constraints online, enabling the system to self-adjust according to the actual operating status, adapt to the operating needs of multiple scenarios in scenic areas, extend the energy storage life, and improve the robustness and adaptability of the control system. Attached Figure Description

[0007] Figure 1 This is a schematic diagram of the overall structure of the present invention; Figure 2 This is a schematic diagram of the rule-based reasoning and preliminary decision-making process of the present invention; Figure 3 This is a schematic diagram of the system state identification and response process of the present invention; Figure 4 This is a schematic diagram of the closed-loop control and adaptive optimization process of the present invention. Detailed Implementation

[0008] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0009] As attached Figure 1 - Appendix Figure 4 The illustrated multi-objective optimization-based DC building microgrid unit control system for scenic areas includes: Status awareness module: used to acquire real-time operating parameters of distributed power sources, energy storage units and loads in DC microgrid; It should be further noted that the state perception module's acquisition frequency is set to 10Hz, and the data transmission delay is ≤50ms.

[0010] It should be specifically noted that the real-time operating parameters acquired by the state perception module include: the DC output power of the photovoltaic array, the output power of the wind turbine, the state of charge of the energy storage battery, the DC bus voltage, the power demand of each load branch, and the time-of-use electricity price signal of the power grid.

[0011] It should be further explained that the DC output power P of the photovoltaic array pv A Hall-effect DC power sensor (model CS100D) is installed at the output of the photovoltaic array combiner box to collect data in real time. The sensor has a range of 0-30kW and an accuracy of ±0.5%FS. The data acquisition method involves directly connecting the sensor in series with the DC circuit. The power signal is converted into a digital signal through electromagnetic induction and then converted into a digital signal by an A / D converter with a sampling accuracy of 16 bits. The output power P of the wind turbine wind The real-time output power is collected by installing a DC power transmitter at the output end of the wind turbine. The transmitter is model BD-48, with a range of 0-20kW and an accuracy of ±0.8%FS. The acquisition method is to connect the transmitter in parallel to the output circuit, monitor the voltage and current, and then calculate the power value. The state of charge (SOC) of the energy storage battery is calculated using the ampere-hour integration method combined with open-circuit voltage calibration. This is achieved by real-time acquisition of the charging and discharging current I using a DC ammeter installed in the main circuit of the energy storage battery pack. bat The ammeter model is ACS712, with a range of 0-50A, an accuracy of ±1%FS, an integration time interval consistent with the acquisition frequency of the state sensing module, and an initial SOC value set to 80%. The formula for calculating the state of charge of an energy storage battery is as follows: Where SOC(0) is the initial state of charge, η bat For charging and discharging efficiency, a value of 0.95 is used during charging and 0.92 is used during discharging. Cn The battery has a rated capacity of 100kWh, and the system voltage is 48V. Every 2 hours, the open-circuit voltage U of the battery pack is collected. oc Calibration is performed, U oc The correspondence with SOC was obtained through experimental fitting, using a three-step core process: controlled charging and discharging + static pressure measurement + data fitting. The specific process is as follows: First, select lithium iron phosphate battery packs of the same model and batch as the system, with 16 cells connected in series to form a 51.2V module, and pair it with high-precision charging and discharging equipment, such as the CT-4008 battery testing system with current accuracy ±0.1%FS, voltage accuracy ±0.05%FS and open circuit voltage tester with resolution 0.01V. The experimental environment is controlled at 25℃±2℃ to avoid the influence of temperature on voltage. First, charge the battery pack at a constant current of 0.2C (approximately 41.7A) until it reaches 100% SOC and a voltage of 54V. Let it rest for 2 hours and record the voltage U at this point. oc1 With SOC1 (100%); Then discharge at a constant current of 0.2C, each time discharging until the SOC decreases by 5%, such as from 100%→95%→90%…→20%. After each discharge, let it stand for 2 hours until the voltage stabilizes, and record the corresponding U. oc Along with the SOC value, a total of 17 sets of core data were collected, with SOC ranging from 20% to 100% and a step size of 5%. The data focused on covering the actual SOC range of the system (20% to 90%) to ensure that the fitting formula was adapted to the system operating conditions. The 17 groups of SOC, U collected oc Data import and data processing tools, due to the U of lithium iron phosphate batteries oc The relationship with SOC is non-linear. A quadratic function model (common and with the highest good fit) was chosen for regression analysis. The final coefficients were: quadratic term coefficient 0.85, linear term coefficient 0.12, and constant term coefficient 0.19, with a goodness-of-fit R-value. 2 A value ≥0.99 indicates an extremely high degree of data-model matching, with an error ≤1%. oc Convert to dimensionless quantity using linear normalization method ,in , That is, we get the formula. .

[0012] DC bus voltage U bus A high-precision DC voltage sensor, model LV25-P, with a range of 0-100V and an accuracy of ±0.2%FS, is installed on the main DC bus to directly acquire the bus voltage. Power demand P of each load branch load-iThe real-time power demand of each branch is collected by installing a micro DC power sensor at the input end of each load branch. The model is DPS300, the measuring range is 0 - 5kW, and the accuracy is ±1%FS. Grid time-of-use electricity price signal C grid (t) Accesses the local power grid marketing system through a 4G module to obtain real-time time-of-use electricity price data. The electricity price is divided into peak periods (9:00 - 12:00, 17:00 - 20:00), flat periods (7:00 - 9:00, 12:00 - 17:00, 20:00 - 22:00), and valley periods (22:00 - 7:00 the next day). The data update frequency is 1 hour / time.

[0013] It should be further noted that the collected original data is filtered using a moving average filter method with a window size of 5 sampling points to remove outliers caused by electromagnetic interference. Through the Modbus-RTU communication protocol, it is transmitted via the RS485 bus to the industrial control computer of the collaborative decision-making module. The CPU model is Intel Core i5-10400, the memory is 8GB, the storage capacity is 512GB SSD, the transmission baud rate is set to 9600bps, and the data verification method is even parity.

[0014] Collaborative decision-making module: includes: Optimization model unit, which pre-sets a multi-objective optimization function including economic operation, energy efficiency, and power supply reliability objectives, as well as the corresponding constraint conditions for the multi-objective optimization function; Rule reasoning unit, which pre-sets a fast response rule base based on system state identification; Among them, the collaborative decision-making module is configured as follows: the rule reasoning unit identifies the current operating state and outputs a preliminary device action combination; then, the preliminary device action combination is used as an initial solution and input into the optimization model unit for iterative optimization to solve the non-dominated solution set that meets the multi-objective optimization function and constraint conditions; finally, a final control strategy is selected from the non-dominated solution set based on a preset decision criterion. It should be specifically noted that the fast response rule base is defined based on the joint state of the system net power and the energy storage state of charge. The joint state at least includes: source-side power enrichment state, load-side power deficit state, and source-load bilateral power antagonism state.

[0015] It should be further noted that when P net ≥3kW and SOC≤85%, it is determined as source-side power enrichment; when P net ≤-2kW and SOC≥25%, it is determined as load-side power deficit; when -2kW < P net <3kW and 25% < SOC < 85%, it is determined as source-load bilateral power antagonism.

[0016] It should be specifically noted that the rule reasoning unit is configured as follows: When the source power enrichment state is identified, the initial equipment action combination output includes: closing the renewable energy generation unit contactor and setting the energy storage converter to charging mode; When a load-side power deficit is identified, the initial equipment action sequence output includes: setting the energy storage converter to discharge mode and generating a non-critical load shedding sequence; When a power antagonism state is identified on both the source and load sides, the initial equipment action combination output includes: maintaining the contactor of the renewable energy generation unit closed and setting the energy storage converter to standby mode.

[0017] It should be further explained that the core of the rule-based reasoning unit is the fast-response rule base, based on the system's net power P. net The joint state definition rules with the energy storage state of charge (SOC) enable rapid identification of the system's operating state and preliminary action output, with a decision delay of ≤100ms; system net power , where P pv P represents the DC output power of the photovoltaic array. wind P represents the output power of the wind turbine. load-i For the power requirements of each load branch, the system net power P net A positive value indicates a power surplus at the source end, a negative value indicates a power deficit at the load end, and a zero value indicates a power balance between the source and the load.

[0018] It should be further explained that the output action combination of the source-side power enrichment state rule inference unit is to close the photovoltaic array and wind turbine contactor + energy storage converter and set them to constant current charging mode; the action execution priority is to first close the renewable energy contactor, and then start charging the energy storage converter after a 200ms delay to avoid current surge; the wind turbine contactor is a CJX2-1210 DC contactor with a rated voltage of 48V and a rated current of 30A, and the energy storage converter is a PCS-50kW with a rated power of 50kW and a conversion efficiency of ≥96%; The load-side power deficit state rule reasoning unit outputs the action combination of setting the energy storage converter to constant power discharge mode and generating a non-critical load shedding sequence. The non-critical load shedding sequence is based on load priority: first, shedding charging piles (lowest priority); if the deficit is still not compensated (P... net ≤-1kW), then cut off general lighting, and finally cut off air conditioning (the highest priority non-critical load), with a cut-off interval of 500ms to avoid a sudden drop in bus voltage; The output action combination of the source-load dual-side power antagonism state rule reasoning unit is to maintain the photovoltaic array and wind turbine contactor closed + the energy storage converter in standby state (DC side voltage tracks bus voltage, AC side is disconnected from grid); in standby state, the energy storage converter response delay is ≤50ms, and it can quickly switch to charging and discharging mode.

[0019] It should be specifically noted that the multi-objective optimization function includes the following objective terms: system operation economics objective term, renewable energy local consumption rate objective term, and energy storage unit lifespan reduction objective term; The constraints include: DC bus voltage stability constraints, energy storage state of charge safety constraints, and load power supply priority constraints.

[0020] It should be further explained that the three objective terms in the multi-objective optimization function are transformed into a single-objective optimization problem using a weighted summation method, and the function expression is as follows: Where ω1, ω2, and ω3 are the dynamic weights of the three objective terms, F1 is the economic target for system operation, F2 is the target for local consumption rate of renewable energy, and F3 is the target for energy storage unit lifespan reduction.

[0021] It should be further explained that the system operation economic target F1 reflects the comprehensive operating cost of the system per unit time, including the grid power purchase cost, energy storage unit depreciation cost, and operation and maintenance cost; the calculation method is as follows: , where P grid The power purchased from the grid is positive when there is a power deficit and zero when there is a surplus. Calculate, C bat C represents the total purchase cost of the energy storage unit. cycle P represents the cycle life of the energy storage unit. bat The charging and discharging power of the energy storage unit is positive during charging and negative during discharging; t is the calculation time step, and C is the value of C. op The system's hourly operation and maintenance cost is a fixed value of 0.5 yuan / hour, which includes the maintenance costs of equipment such as sensors and contactors; Then, F1 is converted into a dimensionless quantity F'1 using the extreme value normalization method: ,in The maximum unit time operating cost preset for the system. The minimum unit time operating cost preset for the system, after standardization .

[0022] The local renewable energy consumption rate target F2 reflects the utilization efficiency of renewable energy, and the goal is to maximize the consumption rate, i.e., minimize the local renewable energy consumption rate. , where η rec To increase the local consumption rate of renewable energy, , where Pabd For renewable energy curtailment, when the energy storage unit is fully charged with a SOC ≥ 90% and the load demand is saturated... Otherwise P abd =0; The energy storage unit lifetime reduction target F3 reflects the aging rate of the energy storage unit. Based on the depth of charge and discharge calculation, the target is to minimize lifetime reduction. The specific calculation formula is as follows: Where DOD represents the depth of charge / discharge of the energy storage unit. DOD max For the energy storage unit, the maximum allowable depth of charge and discharge is 80%, and the SOC is... in The initial state of charge (SOC) for a single charge-discharge cycle. fin This represents the state of charge at the end of a single charge-discharge cycle.

[0023] It should be further explained that the initial values ​​of ω1, ω2, and ω3 are taken into account the common operating scenarios of DC building microgrids in scenic areas: In normal operating scenarios, without extreme loads or energy fluctuations: This is suitable for daily visitor flow in scenic areas (such as weekdays and non-holidays). At this time, the demands for economic efficiency, absorption rate, and energy storage life are relatively balanced. The initial value focuses on a moderate balance. ω1=0.35 Electricity prices are stable during normal periods, and cost control is a medium priority. ω2=0.35 Photovoltaic / wind power output is stable during normal periods, and a basic absorption rate needs to be guaranteed to reduce curtailment. ω3=0.3 Normal charging and discharging depth is moderate, and the risk of equipment lifespan loss is low. Its weight is slightly lower than the previous two. In peak tourist season scenarios, high load and high power demand apply: This is suitable for holidays and peak tourist seasons, where critical load power supply reliability is prioritized, indirectly affecting energy storage lifespan and absorption rate. Initial values ​​emphasize energy storage protection and absorption guarantee. ω1=0.25 During peak season, power supply must be prioritized, temporarily reducing cost control priority; ω2=0.4 During peak season, load demand is high, and a high absorption rate can reduce grid power purchases while avoiding power waste; ω3=0.35 During peak season, energy storage charging and discharging frequencies are high, requiring priority control of lifespan degradation to prevent power outages due to energy storage failures. Off-season cost control scenario, low load + large electricity price fluctuation: Applicable to the off-season of scenic spots (such as winter when there are few tourists and non-tourist periods). At this time, the load demand is low, the difference in time-of-use electricity prices of the grid is significant, and the peak-valley price difference is large. The initial value focuses on economic optimization. ω1=0.45 The off-season revenue is low, and it is necessary to reduce operating costs by optimizing power purchase / energy storage scheduling, which has the highest priority. ω2=0.3 The off-season load is low, and if the output of renewable energy is excessive, it can be moderately abandoned to avoid overcharging of energy storage. ω3=0.25 The off-season energy storage charging and discharging frequency is low, and the risk of lifespan loss is small, so the weight can be appropriately reduced. High-output renewable energy scenarios, high sunshine / high wind speed + low curtailment demand: Applicable to spring and autumn, when sunshine is abundant and wind is stable. At this time, the output of photovoltaic / wind power far exceeds the load demand, so it is necessary to prioritize avoiding curtailment and respond to green and low-carbon policies. The initial value focuses on optimizing the absorption rate. ω1=0.25 When renewable energy output is high, the grid's electricity purchase demand is low, and the cost pressure is small. ω2=0.5 Prioritize the absorption of excess renewable energy through energy storage charging and non-critical load peak-shifting to reduce the curtailment rate. ω3=0.25 During short-term high-output periods, the charging depth can be appropriately increased. As long as the SOC≤90% constraint is not exceeded, the risk of lifespan reduction is controllable. It should be further noted that after setting the initial weight values, they need to be substituted into the multi-objective optimization function for basic constraint verification. If the optimization result under the initial values ​​satisfies the DC bus voltage stability requirement of 45V ≤ U... bus If the voltage is ≤52V and the energy storage SOC safety is 20%≤SOC≤90%, then the initial value is valid. If a constraint violation occurs and ω2 is too high, causing the energy storage to overcharge to SOC>90%, the initial values ​​ω2 and ω3 need to be finely adjusted to decrease by 0.05 and increase by 0.05 until the basic constraints are met.

[0024] It should be further noted that the DC bus voltage stability constraint is: 45V≤U bus ≤52V, when U bus When the voltage is below 45V, the energy storage discharge power should be forcibly reduced or non-critical loads should be cut off; when U bus When the voltage is >52V, force a reduction in energy storage charging power or increase the input of non-critical loads; Energy storage state of charge safety constraints: 20%≤SOC≤90%, achieved by limiting charging and discharging power. When charging, if SOC≥85%, the charging power is gradually reduced; when discharging, if SOC≤25%, the discharging power is gradually reduced. Load power supply priority constraints: the power supply interruption time of critical loads is ≤0.1s, and the power supply interruption time of non-critical loads is ≤1s. This is achieved through the timing arrangement of control commands. The switching command priority of critical loads is higher than that of non-critical loads.

[0025] It should be specifically noted that the preset decision-making criteria are as follows: The dynamic weights of each objective term in the multi-objective optimization function are calculated based on the entropy weight method, and the solution with the highest comprehensive evaluation value is selected from the non-dominated solution set as the final control strategy based on these weights.

[0026] It should be further explained that the non-dominated solution set and the final control strategy selection are solved using a non-dominated sorting genetic algorithm to solve the multi-objective optimization function. The algorithm parameters are set as follows: population size 100, number of iterations 50, crossover probability 0.8, mutation probability 0.05. The algorithm outputs a non-dominated solution set that satisfies all constraints, containing 20 to 30 feasible solutions. The dynamic weights of each objective term are calculated using the entropy weight method, and the steps are as follows: First, the raw data for each objective item are standardized to eliminate the influence of dimensions; then, the information entropy of each objective item is calculated. k = 1 / ln(m), where m is the number of non-dominated solutions, p ij The standardized value proportion of the j-th solution for the i-th objective term; finally, the weights are calculated. This ensures that the weights are dynamically adjusted according to the system's operating status; finally, the comprehensive evaluation value of each non-dominated solution is calculated based on the dynamic weights. The solution with the highest comprehensive evaluation value is selected as the final control strategy, where F ij This is the standardized value of the j-th solution for the i-th objective term.

[0027] It should be specifically noted that the final control strategy is a fine-tuning of the initial equipment action combination. The calibration includes smoothing and optimizing the charging and discharging power curves of the energy storage unit, and dynamically arranging the switching sequence of loads with different priorities.

[0028] It should be further explained that the charging and discharging power curves of the energy storage unit have been optimized for smoothness: a trapezoidal charging curve is adopted to avoid sudden power changes. During charging, the initial 10 minutes are charged with a current of 0.5C, after the SOC reaches 70%, the current is charged with a current of 0.3C, and after the SOC reaches 85%, the current is trickle charged with a current of 0.1C. During discharging, the initial 10 minutes are discharged with a current of 0.4C, and after the SOC drops to 40%, the current is discharged with a current of 0.2C. Dynamically schedule the switching sequence of loads with different priorities: adjust the switching interval based on bus voltage fluctuations, when U bus When the deviation from the target value (48V) is ≤0.5V, the switching interval is 300ms; when U bus When the deviation is between 0.5 and 1V, the switching interval is extended to 500ms; when U bus When the deviation is greater than 1V, suspend the switching of non-critical loads and prioritize stabilizing the bus voltage.

[0029] Control command generation module: used to parse the final control strategy into timed device control commands, so as to control the scenic area DC building microgrid unit to perform tasks based on the device control commands; It should be further explained that the instructions for the contactor are in the format of device ID + action instruction + execution time, for example, photovoltaic contactor-close-2024-05-20 10:00:00.000, the action instruction includes closing and opening, and the execution time is accurate to milliseconds; Commands for energy storage converters should use the format of device ID + operating mode + power / current parameters + execution time, such as energy storage converter-constant current charging-100A-2024-05-20 10:00:00.200. Operating modes include constant current charging, constant power discharging, and standby. Parameters must be within the rated range of the device. Commands for loads: Use the format of load ID + action command + priority + execution time, for example, charging pile-cut-3-2024-05-20 10:00:01.000. The action command includes activation and cut-off, and the priority is 1-3, with 1 being the highest and 3 being the lowest.

[0030] Control commands are sequenced according to the principle of safety priority and step-by-step execution: commands related to bus voltage stability (such as energy storage converter mode switching) are executed first, with an execution delay of ≤50ms; renewable energy access / disconnection commands are executed second, with an execution delay of ≤200ms; non-critical load switching commands are executed last, with an execution delay of ≤500ms; the execution interval of the same type of command is ≥100ms to avoid command conflicts that may cause equipment malfunctions.

[0031] It should be further explained that the control commands are transmitted to each execution unit via the CAN bus at a baud rate of 250kbps. The CRC-16 check method is used to ensure the integrity of the commands. After each command is sent, the receiving unit must return an acknowledgment signal within 50ms. If no acknowledgment signal is received, the control command generation module will resend the command. The number of resends will not exceed 3. If no acknowledgment is received, an alarm signal will be triggered.

[0032] Execution and feedback module: used to execute the control commands and collect system response data to update the input of the state perception module, forming a closed-loop control.

[0033] It should be specifically noted that the system response data monitored by the execution and feedback module includes: The system response data includes the duration of power outages for critical loads, the single-cycle depth of energy storage units, and the instantaneous power curtailment of renewable energy sources. This data is used to adaptively correct the rule thresholds of the rule inference unit and the constraints of the optimization model unit online.

[0034] It should be further explained that the contactor actuator unit configures one CJX2-1210 DC contactor for each of the photovoltaic array and wind turbine generators, installed in the distributed power combiner box, and communicates with the control command generation module via CAN bus; the energy storage converter actuator unit adopts a PCS-50kW bidirectional energy storage converter, supports seamless switching between charging and discharging modes, has a response time ≤50ms, is installed in the energy storage cabinet, and communicates with the control command generation module via Modbus-TCP protocol; the load control actuator unit configures one miniature DC circuit breaker (model DZ47-63DC, rated voltage 48V, rated current 10A) for each load branch, controlled by a relay module (model G2R-1-E, rated voltage 48V), installed in the load distribution box, and communicates with the control command generation module via RS485 bus.

[0035] It should be further noted that the execution and feedback module collects system response data through an independent sensor network at a frequency of 5Hz: critical load power interruption duration T int A voltage monitor, model VM-01, with an accuracy of ±0.1V, is installed at the critical load input. Timing begins when the voltage drops below 45V and stops when it recovers above 45V. int This is the timing difference; Energy storage unit single cycle depth of DOD cyc The SOC values ​​are calculated by collecting data before and after a single charge-discharge cycle. SOC pre State of charge (SOC) before a single charge and discharge cycle suf The state of charge after a single charge and discharge cycle; Instantaneous curtailment of renewable energy P abd A power sensor is installed at the outlet of the renewable energy combiner box to collect the actual power P connected to the system. acc , .

[0036] It should be further noted that the system response data is transmitted to the collaborative decision-making module in real time, and online adaptive correction is performed once per hour. Rule inference unit rule threshold correction based on T int and P abd Adjust the joint state threshold if T int >0.05s indicates insufficient response under load-side power deficit conditions. The P value under load-side power deficit conditions should be adjusted accordingly. net The threshold was adjusted from -2kW to -1.8kW; if P abd >3kW indicates that the energy storage charging power is insufficient under the source-end power enrichment state. The P in the source-end power enrichment state... net The threshold was adjusted from 3kW to 2.8kW; Optimization of model unit constraint correction based on DOD cyc Adjusting the safety constraints of the energy storage state of charge (DOD) if there are 3 consecutive DOD events. cyc If the SOC (State of Charge) is >70%, it indicates that the energy storage unit's cycle depth is too large. The SOC safety upper limit will be adjusted from 90% to 88%, and the lower limit from 20% to 22%. If there are three consecutive DOD (Demand of Activation) cycles... cyc If the efficiency is less than 30%, it indicates that the energy storage unit is not efficient enough. The upper limit of SOC safety is adjusted to 92%, and the lower limit is adjusted to 18%.

[0037] The following example, using the weekday (9:00-10:00, peak electricity pricing) operation of a visitor center in a mountain scenic area as an example, details the overall system operation process: The state awareness module collected the following parameters: P pv =15kW, light intensity 900lux, P wind =8kW, wind speed 7m / s, SOC=60%, U bus =47.8V, Critical load 6kW, non-critical load 10kW The data is filtered and then transmitted to the collaborative decision-making module. calculate If Pnet≥3kW and SOC≤85%, it is determined to be in a power enrichment state at the source end. The rule reasoning unit outputs the preliminary equipment action combination: close the photovoltaic and wind turbine contactors + constant current charging of 100A for the energy storage converter.

[0038] Using the initial action combination as the initial solution, the multi-objective optimization function is substituted, and the dynamic weights ω1=0.4, ω2=0.35, and ω3=0.25 are calculated using the entropy weight method. After obtaining the non-dominated solution set, the solution with the highest comprehensive evaluation value is selected as the final control strategy: closing the photovoltaic and wind turbine contactors + constant current charging (120A, corresponding to a charging power of 5.76kW) of the energy storage converter + investing in one charging pile (2kW). This strategy improves the renewable energy consumption rate while ensuring economic efficiency, and the energy storage charging power is within a safe range.

[0039] The control instruction generation module parses the final control strategy and generates timing-based instructions: 9:00:01.150: Photovoltaic contactor - closed - 2024-05-20 09:00:01.150; Wind turbine contactor - closed - 2024-05-20 09:00:01.250; 9:00:01.350: Energy Storage Converter - Constant Current Charging - 120A - 2024-05-20 09:00:01.350; 9:00:01.650: Charging pile 1-Industry-3-2024-05-20 09:00:01.650.

[0040] Each execution unit receives and executes instructions, resulting in the contactor closing, the energy storage converter starting charging, and the charging pile being put into operation.

[0041] The execution and feedback module collects system response data: T int =0.02s, critical load showed no significant interruption, DOD cyc =15%, after a single charge, the SOC increases from 60% to 75%, P abd =0kW, renewable energy is fully consumed. The data is transmitted to the status perception module to update the input. At the same time, the collaborative decision-making module verifies the rule thresholds and constraints based on the data. Since all indicators are within a reasonable range, no adjustments are made.

[0042] In this embodiment, the core performance indicators of the system are as follows: The local consumption rate of renewable energy is ≥95% under normal weather conditions; the system operating cost is reduced by 15% to 20% compared with traditional DC microgrids; the cycle life of energy storage units is extended by 10% to 15% based on DOD optimization; the power supply reliability of critical loads is ≥99.99%, and the annual interruption time is ≤5 minutes; All power equipment uses DC-specific models with overvoltage, overcurrent, and overtemperature protection; sensors are industrial-grade products with a protection level of ≥IP65, suitable for harsh outdoor environments in scenic areas; communication lines use shielded cables and are equipped with surge protectors, model SPD-48DC, to prevent lightning interference.

[0043] The collaborative decision-making module adopts a dual-machine hot standby architecture, with a master-slave switching time of ≤1s to avoid single point of failure; control commands are transmitted in encrypted form, with the key updated every hour to prevent commands from being tampered with; the system has a self-diagnostic function, which checks the operating status of each module every 5 minutes, and immediately triggers an audible and visual alarm and records the fault log when a fault is detected.

[0044] Secondly: The accompanying drawings of the embodiments disclosed in this invention only involve the structures involved in the embodiments disclosed in this invention. Other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of this invention can be combined with each other. In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A DC building microgrid unit control system based on multi-objective optimization for scenic spots, characterized in that, include: Status awareness module: used to acquire real-time operating parameters of distributed power sources, energy storage units and loads in DC microgrid; The collaborative decision-making module includes: The optimization model unit is pre-set with a multi-objective optimization function that includes objectives of economic operation, energy efficiency and power supply reliability, as well as the multi-objective optimization function and its corresponding constraints. The rule reasoning unit is pre-loaded with a fast-response rule base based on system state identification; The collaborative decision-making module is configured to: identify the current operating state by the rule reasoning unit and output a preliminary combination of equipment actions; input the preliminary combination of equipment actions as the initial solution into the optimization model unit for iterative optimization to solve for the non-dominated solution set that satisfies the multi-objective optimization function and constraints; and select a final control strategy from the non-dominated solution set based on a preset decision criterion. Control command generation module: used to parse the final control strategy into timed device control commands, so as to control the scenic area DC building microgrid unit to perform tasks based on the device control commands; Execution and feedback module: used to execute the control commands and collect system response data to update the input of the state perception module, so as to form closed-loop control.

2. The DC building microgrid unit control system based on multi-objective optimization for scenic spots according to claim 1, characterized in that, The real-time operating parameters acquired by the state perception module include: the DC output power of the photovoltaic array, the output power of the wind turbine, the state of charge of the energy storage battery, the DC bus voltage, the power demand of each load branch, and the time-of-use electricity price signal of the power grid.

3. The scenic area DC building microgrid unit control system based on multi-objective optimization according to claim 1, characterized in that, The objective terms of the multi-objective optimization function include: the system operation economics objective term, the local renewable energy consumption rate objective term, and the energy storage unit lifetime loss objective term. The constraints include: DC bus voltage stability constraints, energy storage state of charge safety constraints, and load power supply priority constraints.

4. A scenic area DC building microgrid unit control system based on multi-objective optimization according to claim 1, characterized in that, The fast response rule base defines rules based on the combined state of system net power and energy storage state of charge. The combined states include at least: source-side power enrichment state, load-side power deficit state, and source-load dual-side power antagonism state.

5. A multi-objective optimization-based DC building microgrid unit control system for scenic areas according to claim 1, characterized in that, The rule reasoning unit is configured as follows: When the source power enrichment state is identified, the initial equipment action combination output includes: closing the renewable energy generation unit contactor and setting the energy storage converter to charging mode; When a load-side power deficit is identified, the initial equipment action sequence output includes: setting the energy storage converter to discharge mode and generating a non-critical load shedding sequence. When a power antagonism state is identified on both the source and load sides, the initial equipment action combination output includes: maintaining the contactor of the renewable energy generation unit closed, and setting the energy storage converter to standby mode.

6. A scenic area DC building microgrid unit control system based on multi-objective optimization according to claim 1, characterized in that, The preset decision criteria are: The dynamic weights of each objective term in the multi-objective optimization function are calculated based on the entropy weight method, and the solution with the highest comprehensive evaluation value is selected from the non-dominated solution set according to the weights as the final control strategy.

7. A scenic area DC building microgrid unit control system based on multi-objective optimization according to claim 1, characterized in that, The final control strategy is a fine-tuning of the initial equipment action combination. The calibration includes smoothing and optimizing the charging and discharging power curves of the energy storage unit, and dynamically arranging the switching sequence of loads with different priorities.

8. A scenic area DC building microgrid unit control system based on multi-objective optimization according to claim 1, characterized in that, The system response data monitored by the execution and feedback module includes: The system response data includes the duration of power outages for critical loads, the single-cycle depth of energy storage units, and the instantaneous power curtailment of renewable energy sources. This data is used to adaptively correct the rule thresholds of the rule inference unit and the constraints of the optimization model unit online.

Images