Operation optimization scheduling method of expressway tunnel micro-storage system

Abstract

The invention provides an operation optimization scheduling method of an expressway tunnel micro-storage system. The system is composed of a photovoltaic power generation unit, a wind power generation unit, a diesel power generation unit, a storage battery, a super-capacitor energy storage unit and a power grid interaction interface. The method comprises the following steps: acquiring system inherent parameters, sun position, weather and traffic flow prediction data; calculating wind power and ideal photovoltaic output; aiming at shadow influence of a mountain in a tunnel environment, dynamically calculating a tunnel shading loss coefficient based on a geometric projection principle, and correcting photovoltaic output; calculating tunnel dynamic electrical load according to traffic flow prediction; establishing a multi-objective optimization model and solving by taking system power balance and equipment constraint as conditions and taking operation cost, expected power shortage and carbon emission minimization as objectives to obtain a day-ahead optimization scheduling scheme of each unit and energy storage; according to the invention, the photovoltaic output prediction precision in the tunnel environment and the reliability, economy and environmental protection of the scheduling scheme are improved.

Description

Technical Field

[0001] This invention relates to the field of power system operation and control technology, specifically to an operation optimization and scheduling method for micro-storage systems in highway tunnels. Background Technology

[0002] As crucial transportation infrastructure, highway tunnels require a stable and reliable power supply around the clock. Maintaining essential power supply capacity is especially vital during grid failures and system islanding. Currently, to improve system reliability and economy, some highway tunnels have been equipped with micro-storage systems incorporating distributed renewable energy sources such as photovoltaics and wind power. However, the photovoltaic power output of highway tunnels is affected by various environmental factors, exhibiting significant uncertainty and volatility, posing challenges to system scheduling and optimization.

[0003] When existing photovoltaic (PV) power output prediction models are used to predict PV power output in highway tunnels, the impact of terrain and structures on the PV array output is often ignored, resulting in a significant deviation between the predicted PV power output curve and the actual value. This deviation will further propagate to subsequent load energy matching, energy storage scheduling, and operation optimization, causing the scheduling strategy to deviate from the actual operating conditions, reducing the system's economic efficiency, reliability, and environmental benefits, and making it difficult to meet the efficient and accurate scheduling requirements of highway tunnel power supply systems. Summary of the Invention

[0004] This application provides an optimized scheduling method for a micro-storage system in a highway tunnel. The tunnel micro-storage system includes a photovoltaic power generation unit, a wind power generation unit, a diesel power generation unit, a battery energy storage unit, a supercapacitor energy storage unit, and a grid interface. The photovoltaic power generation unit includes a photovoltaic array. The photovoltaic array is constructed along the tunnel entrance direction. The method includes: Obtain the inherent parameters of the micro-storage system in the highway tunnel during the scheduling cycle; the inherent parameters of the system include the height of the mountain above the tunnel, the size and orientation of the photovoltaic array, the azimuth angle of the tunnel, the power characteristic curve of the wind turbine, and the power parameters of the photovoltaic array; Obtain the solar altitude angle and solar azimuth angle at each scheduling moment within the scheduling cycle; Obtain meteorological forecast data for each scheduling time within the scheduling cycle of the tunnel location; the meteorological forecast data includes solar irradiance, wind direction, and wind speed; Obtain traffic flow prediction data for each scheduling time within the scheduling cycle of the tunnel location; Based on wind direction, wind speed, and the power characteristic curves of wind turbines, the theoretical wind power output of the wind power generation unit at each scheduling moment within the scheduling cycle is calculated; based on solar irradiance and the power parameters of the photovoltaic array, the ideal photovoltaic output of the photovoltaic power generation unit at each scheduling moment within the scheduling cycle is calculated. Calculate the tunnel shading loss coefficient at each scheduling time within the scheduling cycle; the tunnel shading loss coefficient is used to characterize the degree to which the photovoltaic output of the photovoltaic array is reduced due to the shadow of the mountain. For scheduling time t within the scheduling cycle, the ideal photovoltaic output at scheduling time t is corrected using the tunnel shading loss coefficient at scheduling time t to obtain the theoretical photovoltaic output at scheduling time t, thereby obtaining the theoretical photovoltaic output at each scheduling time within the scheduling cycle. Based on traffic flow prediction data and system inherent parameters, the total electrical load of the tunnel at each scheduling time within the scheduling cycle is calculated. Based on the theoretical photovoltaic output, theoretical wind power output, and total tunnel electrical load at each scheduling moment within the scheduling cycle, and with system power balance and equipment operation constraints as constraints, and with the optimization objectives of minimizing total system operating cost, minimizing expected power shortage, and minimizing total system carbon emissions, an operation optimization model for a highway tunnel micro-storage system is established. The decision variables of the model include the battery charging power, battery discharging power, supercapacitor charging power, supercapacitor discharging power, diesel generator output power, wind curtailment power, solar curtailment power, and the power exchanged with the grid through the grid interface at each scheduling moment within the scheduling cycle. The model is solved to obtain the final runtime scheduling scheme; The calculation of the tunnel shading loss coefficient at each scheduling time within the scheduling cycle includes: For scheduling time t within the scheduling cycle, based on the height of the mountain above the tunnel, the size and orientation of the photovoltaic array, the azimuth angle of the tunnel, the solar altitude angle and solar azimuth angle at scheduling time t, the ratio of the area of ​​the photovoltaic array covered by the mountain shadow to the total area of ​​the photovoltaic array at scheduling time t is calculated using the principle of geometric projection. Based on the ratio of the area of ​​the photovoltaic array covered by shadow to the total area of ​​the photovoltaic array at scheduling time t, the tunnel shading loss coefficient at scheduling time t is determined, so as to obtain the tunnel shading loss coefficient at each scheduling time within the scheduling cycle.

[0005] Furthermore, for scheduling time t within the scheduling cycle, based on the height of the mountain above the tunnel, the size and orientation of the photovoltaic array, the azimuth angle of the tunnel, and the solar altitude and azimuth angles at scheduling time t, the ratio of the area of ​​the photovoltaic array covered by the mountain shadow to the total area of ​​the photovoltaic array at scheduling time t is calculated using the principle of geometric projection, including: Calculate the ratio of the area of ​​the photovoltaic array covered by the mountain's shadow to the total area of ​​the photovoltaic array at scheduling time t according to Formula 1; Formula 1; Among them, A blo (t) represents the area of ​​the photovoltaic array covered by shadow at scheduling time t, A tol (t) Total area of ​​the photovoltaic array, L pvL is the length of the photovoltaic array along the tunnel entrance, and d is the horizontal distance from the foot of the mountain to the near edge of the photovoltaic array; shade Let t be the length of the shadow of the mountain from its foot to the end of the shadow that is far from the foot of the mountain; ; Let be the solar azimuth angle at time t. The azimuth is the clockwise angle between the tunnel exit direction and due south. h(t) is the solar altitude angle; H is the height of the mountain. The tunnel shading loss coefficient at scheduling time t is determined based on the ratio of the area of ​​the photovoltaic array covered by shadow to the total area of ​​the photovoltaic array, including: The occlusion ratio K at scheduling time t is calculated using Formula 2. shade (t); Formula 2; The tunnel shading loss coefficient k at scheduling time t is calculated according to Formula 3. Formula 3.

[0006] Furthermore, the meteorological forecast data also includes ambient temperature, and the system's inherent parameters include the nominal operating temperature of the photovoltaic modules; the methods also include: The actual temperature of the photovoltaic array at scheduling time t within the scheduling cycle is calculated according to Formula 4. Formula 4; Among them, T c (t) represents the actual temperature of the photovoltaic array at scheduling time t; T e G(t) represents the ambient temperature at scheduling time t, G(t) represents the solar irradiance at scheduling time t, and NOCT represents the nominal cell operating temperature of the photovoltaic module. The product of the difference between the actual temperature of the photovoltaic array at scheduling time t and the temperature of the photovoltaic array under standard test conditions and the preset temperature coefficient is used as the temperature correction coefficient at scheduling time t. The ideal photovoltaic output at scheduling time t is corrected using the tunnel shading loss coefficient at scheduling time t, and the theoretical photovoltaic output at scheduling time t is obtained. This includes correcting the ideal photovoltaic output at scheduling time t using the tunnel shading loss coefficient at scheduling time t and the temperature correction coefficient at scheduling time t, and obtaining the theoretical photovoltaic output at scheduling time t.

[0007] Furthermore, the system's inherent parameters also include the fixed load of the monitoring and safety system; the total electrical load of the tunnel at each scheduling moment within the scheduling cycle is calculated, including: Calculate the lighting load at each scheduling time within the scheduling cycle; the lighting load includes a fixed component and a variable component that is positively correlated with traffic flow forecast data; Calculate the ventilation load at each scheduling time within the scheduling cycle; the ventilation load includes a fixed component and a variable component that is positively correlated with the traffic flow forecast data; The total electrical load of the tunnel at each scheduling time within the scheduling cycle is obtained by adding the lighting load, ventilation load, and monitoring and safety system fixed load at each scheduling time within the scheduling cycle.

[0008] Furthermore, the system's inherent parameters also include fan efficiency, and the ventilation load at each scheduling moment within the scheduling cycle includes: According to Formula 5, calculate the variable part of the ventilation load at scheduling time t within the scheduling cycle; Formula 5; in, (t) is the pressure difference during wind turbine operation at scheduling time t; Q v (t) represents the required ventilation volume of the fan at scheduling time t. For fan efficiency.

[0009] Furthermore, the system's inherent parameters also include air density, friction coefficient along the tunnel, tunnel length, equivalent tunnel diameter, tunnel cross-sectional area, and piston wind coefficient; the ventilation load for each scheduling moment within the scheduling cycle also includes: According to Formula 6, the calculation is as follows: p; Formula 6; in, For air density, λ, L, D h These represent the friction coefficient along the tunnel, tunnel length, and equivalent tunnel diameter, respectively; A is the tunnel cross-sectional area, and k... tr Let Q(t) be the piston wind coefficient, and Q(t) be the traffic flow prediction data at scheduling time t.

[0010] Furthermore, the ventilation load at each scheduling time within the scheduling cycle also includes: calculating the ventilation volume required by the fan at scheduling time t; wherein the ventilation volume required by the fan at scheduling time t is the larger of the ventilation volume required to maintain the air pollutant concentration in the tunnel within the pollutant limit value and the ventilation volume required to maintain the temperature rise in the tunnel within the safety limit value at scheduling time t.

[0011] Furthermore, the inherent parameters of the system also include the pollutant mass emission rate and thermal power of each vehicle, the maximum allowable concentration of pollutants in the tunnel, the specific heat capacity of the air, the air density, and the maximum acceptable temperature rise in the tunnel; meteorological forecast data includes the pollutant concentration of the air outside the tunnel itself during the scheduling cycle; the ventilation volume required by the fan at scheduling time t is calculated according to Formula 7. Formula 7; Among them, e poll The pollutant emission rate per vehicle; C max C0 represents the maximum permissible concentration of pollutants in the tunnel, while h represents the concentration of pollutants in the air outside the tunnel during the scheduling cycle. For air density, c p ΔT is the specific heat capacity of air. max This represents the maximum acceptable temperature rise for the tunnel.

[0012] Furthermore, the methods also include: Obtain the power grid state coefficient at each scheduling moment within the scheduling cycle; when the power grid is faulty, the power grid state coefficient is 1, and when the power grid is operating normally, the power grid state coefficient is 0. Minimizing the expected power shortage EENS is expressed by Equation 8: Formula 8; Where γ(t) is the power grid state coefficient.

[0013] Furthermore, by solving the model, the final runtime scheduling scheme is obtained, including: The NSGA-II algorithm is used to solve the operation optimization model to obtain the Pareto optimal solution set of the trade-off relationship between the total system operating cost, the expected power shortage, and the total carbon emissions of the system. The VIKOR multi-criteria decision method is used to select the optimal compromise solution from the Pareto optimal solution set as the final operation scheduling scheme. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating the operation optimization and scheduling method for a highway tunnel micro-storage system according to an embodiment of this application.

[0015] Figure 2 This is a schematic diagram illustrating the calculation principle of the tunnel shading loss coefficient in an embodiment of this application.

[0016] Figure 3 This is a schematic diagram of the orientation, solar azimuth angle, and tunnel azimuth angle of the photovoltaic array in an embodiment of this application. Detailed Implementation

[0017] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0018] This invention provides an optimized scheduling method 10 for a micro-storage system in a highway tunnel. The tunnel micro-storage system includes a photovoltaic power generation unit, a wind power generation unit, a diesel power generation unit, a battery energy storage unit, a supercapacitor energy storage unit, and a grid interface; the photovoltaic power generation unit includes a photovoltaic array; the photovoltaic array is constructed along the tunnel entrance direction. Specifically, the photovoltaic array is located outside the tunnel exit, constructed along the tunnel exit direction.

[0019] Method 10 is applicable to the daily operation and scheduling phase of the system after its construction, especially suitable for short-cycle (e.g., the next 24 hours) day-ahead optimized scheduling. After acquiring meteorological forecast data and traffic flow forecast data for each scheduling moment within the scheduling cycle, and calculating the wind power, photovoltaic power generation output, and tunnel load for each scheduling moment within the scheduling cycle based on the meteorological forecast data and traffic flow forecast data, Method 10 constructs and solves a multi-objective optimization model to determine the optimized output plan of each controllable device at each scheduling moment (e.g., every 15 minutes or hour is a scheduling step, the scheduling moment is the start time of each scheduling step, and the output and power variables of each component remain unchanged within each scheduling step). This includes: battery charging power, battery discharging power, supercapacitor charging power, supercapacitor discharging power, diesel generator output power, wind curtailment power, solar curtailment power, and power interaction with the grid at each scheduling moment within the scheduling cycle. The optimization objective is to balance the economic efficiency, power supply reliability, and environmental friendliness of the system operation while ensuring reliable power supply to the tunnel.

[0020] Understandably, the number of scheduling moments within each scheduling cycle is determined by the scheduling cycle and the scheduling step size. For example, if a scheduling cycle is 24 hours and the scheduling step size is 1 hour, then the scheduling cycle contains 24 scheduling steps, including scheduling moment 0, scheduling moment 1, ..., scheduling moment 23, for a total of 24 scheduling moments. That is, the scheduling moments are numbered from 0 to T, where T = scheduling cycle / scheduling step size - 1.

[0021] The following combination Figure 1 The embodiments of this application will be described in detail below, including specific steps.

[0022] Method 10 includes: Step 110: Obtain the inherent system parameters of the highway tunnel micro-storage system during the scheduling cycle; the inherent system parameters include the height of the mountain above the tunnel, the size and orientation of the photovoltaic array, the azimuth angle of the tunnel, the power characteristic curve of the wind turbine, and the power parameters of the photovoltaic array.

[0023] This step aims to collect the fixed physical and electrical parameters of various devices and the tunnel environment within the optimized micro-storage system. These parameters are the foundational data for theoretical output calculations, load modeling, and constraint settings, and can be preset or measured.

[0024] The scheduling period refers to the future time period covered by the optimized scheduling, usually taken as 24 hours, to correspond to the day-ahead scheduling needs.

[0025] System-inherent parameters refer to parameters of the equipment and tunnel environment within the system that are determined after design or construction and remain relatively unchanged during the scheduling cycle. System-inherent parameters can include the following categories: Tunnel environmental parameters: The height of the mountain above the tunnel and the azimuth of the tunnel. The height of the mountain above the tunnel can specifically refer to the height of the mountain near the tunnel exit. Some long tunnels can reach several kilometers in length, and the mountain range where the tunnel is located often consists of multiple continuous hills. Hills far from the photovoltaic panels at the tunnel exit will have almost no impact on the photovoltaic array. Therefore, only the height of the mountain above the tunnel exit needs to be considered.

[0026] Photovoltaic power generation unit parameters: Dimensions of the photovoltaic array (such as the length of the photovoltaic array along the tunnel entrance, the total area of ​​the photovoltaic array), orientation (such as the horizontal distance from the foot of the mountain to the near edge of the photovoltaic array), rated power of the photovoltaic modules, temperature coefficient, nominal cell operating temperature (NOCT); the ratio of photovoltaic output when the array is completely shaded by the mountain to photovoltaic output when it is not shaded under the same solar irradiance, etc.

[0027] Wind power generation unit parameters: rated power of wind turbine generator set, cut-in / cut-out wind speed, power characteristic curve parameters, etc.

[0028] Energy storage unit parameters: Rated power (charge / discharge), rated capacity, charge / discharge efficiency, upper and lower limits of state of charge (SOC), initial SOC, operating cost factor, start-stop cost factor, etc. of batteries and supercapacitors.

[0029] Diesel generator unit parameters: Rated power, minimum technical output, gradeability, operating cost coefficient, carbon emission factor, start-up and shutdown cost coefficient, etc.

[0030] Grid interaction parameters: power transmission limit of the line connected to the grid, electricity purchase and sale price (time-of-use price), grid carbon emission factor, etc.

[0031] Tunnel load parameters: reference lighting power, proportion of adjustable lighting and total rated power, ventilation fan base load power, pollutant emission related parameters (such as vehicle emission rate, allowable concentration limit), air characteristic parameters, etc.

[0032] Step 120: Obtain the solar altitude angle and solar azimuth angle at each scheduling moment within the scheduling cycle.

[0033] The solar altitude angle and solar azimuth angle can be calculated from the latitude and longitude of the tunnel location and the date of the scheduling cycle.

[0034] Step 130: Obtain meteorological forecast data for each scheduling time within the scheduling cycle of the tunnel location; the meteorological forecast data includes solar irradiance, wind direction, and wind speed. This step obtains weather condition forecasts for future times within the scheduling cycle, which is a key input for predicting renewable energy output and partial load.

[0035] Meteorological forecast data can be obtained from quantitative data of future weather conditions provided by meteorological forecast services, or by calculation based on quantitative data of future weather conditions provided by meteorological forecast services, or by prediction based on historical meteorological data and / or meteorological data of the past N days using time series analysis or machine learning models (such as LSTM).

[0036] Understandably, weather forecast data is in time series form. For example, if the scheduling period is 24 hours and the scheduling time is each hour within 24 hours, then the weather forecast data includes the ambient temperature, solar irradiance, wind direction, and wind speed for each hour within 24 hours.

[0037] Step 140: Obtain traffic flow prediction data for each scheduling time within the scheduling cycle of the tunnel location.

[0038] This step obtains forecast information on the number of vehicles passing through the tunnel at future times within the scheduling cycle. Traffic flow is the most important factor determining the tunnel's variable load (lighting, ventilation).

[0039] Traffic flow forecast data is usually based on historical traffic flow data, date type (weekday / holiday), weather conditions, special events and other influencing factors, and is obtained by using time series analysis or machine learning models (such as LSTM).

[0040] Step 150: Based on wind direction, wind speed, and the power characteristic curve of the wind turbine, calculate the theoretical wind power output of the wind power generation unit at each scheduling moment within the scheduling cycle; based on solar irradiance and the power parameters of the photovoltaic array, calculate the ideal photovoltaic output of the photovoltaic power generation unit at each scheduling moment within the scheduling cycle.

[0041] This step uses the meteorological forecast data from step 130 and the equipment parameters from step 110 to calculate the maximum possible power that the wind and photovoltaic power generation units can generate at each time without considering any operational limitations or obstructions.

[0042] For example, theoretical wind power output can be calculated based on wind speed, wind direction, and the power characteristic curve of the wind turbine in the system's inherent parameters from meteorological forecast data. Ideal photovoltaic power output can be calculated based on solar irradiance G(t) in meteorological forecast data and the power parameters of the photovoltaic array in the system's inherent parameters (including solar irradiance G under standard test conditions). stcRated power P of photovoltaic array under standard test conditions stc This is calculated. For example, the ideal photovoltaic output is... .

[0043] It is understandable that the theoretical wind power output of the wind power generation unit at each scheduling moment within the scheduling cycle and the ideal photovoltaic power output of the photovoltaic power generation unit at each scheduling moment within the scheduling cycle are in time series form.

[0044] Step 160: Calculate the tunnel shading loss coefficient at each scheduling time within the scheduling cycle; the tunnel shading loss coefficient is used to characterize the degree to which the photovoltaic output of the photovoltaic array is reduced by the shadow of the mountain.

[0045] This is one of the key steps in this invention, aiming to quantify the impact of the shading effect of the mountain on the photovoltaic array on photovoltaic output under the special environment of the tunnel, and to convert it into a loss coefficient between 0 and 1. A coefficient of 1 indicates no effect, while a coefficient less than 1 indicates that the photovoltaic output is reduced due to the shading effect of the mountain.

[0046] In their research on the actual operation optimization and simulation of a micro-storage system in a highway tunnel, the inventors of this application discovered that, under the same meteorological conditions, the theoretical power output prediction based on a general photovoltaic model is significantly higher than the actual observable available photovoltaic power in the tunnel scenario. This discrepancy directly leads to an overestimation of photovoltaic contribution in day-ahead scheduling plans, resulting in increased operating costs, insufficient reserve capacity, and even affecting the power supply reliability during islanded operation.

[0047] To investigate the root cause of this deviation, the inventors conducted in-depth research on the unique spatial structure and terrain environment of tunnel scenarios. In photovoltaic projects at highway tunnel entrances in my country, photovoltaic arrays are often installed on the natural slopes on both sides of the tunnel entrance, laid in the median strip of the highway, or constructed as photovoltaic canopies and deployed above the tunnel entrance, mostly along the highway route. These layouts are adjacent to mountains, and the mountain's shadow easily obstructs the photovoltaic arrays, especially during periods of low solar altitude such as early morning and evening, when the mountain's shadow directly covers part of the photovoltaic panel area. If seasonal changes cause variations in the angle of direct sunlight, the range and duration of shadow coverage will further change, thus affecting the effective light-receiving area and power generation efficiency of the photovoltaic modules.

[0048] Through extensive field research, illumination geometry analysis, and historical data verification, the inventors of this application have discovered that the failure to account for the dynamic shading effect caused by the special environment of the tunnel when calculating photovoltaic output is the main reason for the serious inaccuracy of photovoltaic output prediction in this scenario, which in turn affects the overall optimization scheduling effect.

[0049] To address this issue, this application creatively introduces the physical quantity of tunnel shading loss coefficient and establishes a refined calculation model for it. This coefficient is based on the height of the mountain where the tunnel is located, the size and orientation of the photovoltaic array, and the real-time position of the sun. It dynamically quantifies the output reduction caused by the loss of effective light-receiving area of ​​the array due to shadow coverage through a geometric projection method.

[0050] Specifically, calculating the tunnel shading loss coefficient at each scheduling time within the scheduling cycle includes, for each scheduling time t within the scheduling cycle: First, the shadow area is calculated. Based on the height of the mountain above the tunnel, the size and orientation of the photovoltaic array, the azimuth angle of the tunnel, and the solar altitude angle and solar azimuth angle at scheduling time t, the ratio of the area of ​​the photovoltaic array covered by the mountain shadow to the total area of ​​the photovoltaic array at scheduling time t is calculated using the principle of geometric projection.

[0051] Next, the tunnel shading loss coefficient is calculated. Based on the ratio of the area of ​​the photovoltaic array covered by shadow to the total area of ​​the photovoltaic array at scheduling time t, the tunnel shading loss coefficient at scheduling time t is determined, so as to obtain the tunnel shading loss coefficient at each scheduling time within the scheduling cycle.

[0052] Step 170: For scheduling time t within the scheduling cycle, the ideal photovoltaic output at scheduling time t is corrected using the tunnel shading loss coefficient at scheduling time t to obtain the theoretical photovoltaic output at scheduling time t, thereby obtaining the theoretical photovoltaic output at each scheduling time within the scheduling cycle.

[0053] This step applies the shading effect calculated in step 160 to the ideal photovoltaic output obtained in step 150, thereby obtaining a theoretical photovoltaic output that is more consistent with the actual environment of the tunnel.

[0054] For example, for each scheduling time t, the ideal photovoltaic output at that time is multiplied by the tunnel shading loss factor at that time to obtain the corrected theoretical photovoltaic output. This process is repeated for all scheduling times within the scheduling period to obtain the theoretical photovoltaic output sequence for each scheduling time within the entire scheduling period.

[0055] Step 180: Based on traffic flow prediction data and system inherent parameters, calculate the total tunnel electrical load at each scheduling time within the scheduling cycle.

[0056] The total electrical load of a tunnel includes ventilation load, lighting load, etc. It is understandable that the greater the traffic volume, the greater the total electrical load of the tunnel. The total electrical load of the tunnel at each scheduling time within the scheduling cycle can be calculated based on the traffic volume forecast data for each scheduling time within the scheduling cycle. Step 190: Based on the theoretical photovoltaic output, theoretical wind power output, and total tunnel electrical load at each scheduling moment within the scheduling cycle, and with system power balance and equipment operation constraints as constraints, and with the optimization objectives of minimizing total system operating cost, minimizing expected power shortage, and minimizing total system carbon emissions, establish an operation optimization model for the highway tunnel micro-storage system. The decision variables of the model include the battery charging power, battery discharging power, supercapacitor charging power, supercapacitor discharging power, diesel generator output power, wind curtailment power, solar curtailment power, and the interaction power with the grid through the grid interface at each scheduling moment within the scheduling cycle.

[0057] This step integrates the theoretical wind and solar power output and total tunnel electrical load obtained from the previous steps, and considers all physical and operational constraints of the system to construct a multi-objective optimization model. The solution to this model will provide the optimal scheduling instructions for each controllable device at each scheduling time.

[0058] The unknowns that need to be solved during the optimization of decision variables represent the specific content of the scheduling scheme. Decision variables include: battery charging power, battery discharging power, supercapacitor charging power, supercapacitor discharging power, diesel generator output power, wind curtailment power, solar curtailment power, and power exchanged with the grid through the grid interaction interface (positive for electricity purchase, negative for electricity sale).

[0059] Optimization objectives include: Minimize the total system operating cost. Total operating cost includes the operation and maintenance costs of all equipment, start-up and shutdown costs, and the costs of trading with the grid (electricity purchase expenditure minus electricity sales revenue).

[0060] Minimize the expected power shortage (EENS). This primarily considers the power supply reliability of the system under islanded operation (grid failure) scenarios. Minimizing this value means improving the reliability during islanded operation.

[0061] Minimize the system's total carbon emissions. Consider the CO2 emissions from diesel generator power generation and electricity purchased from the grid (which may originate from thermal power plants). Minimizing this value means improving the system's environmental friendliness.

[0062] One of the optimization objectives of highway tunnel micro-storage systems is to minimize the total operating cost of n tunnels (n ≥ 1) within a scheduling cycle. The total operating cost of a highway tunnel micro-storage system mainly consists of the operating cost of the tunnel micro-storage system, the start-up and shutdown costs of the equipment in the micro-storage system, and the electricity purchase cost and electricity sales revenue of the micro-storage system. In some embodiments, the operating cost of a micro-storage system includes the operating costs of the distributed photovoltaic power generation system, the distributed wind power generation system, the hybrid battery and supercapacitor energy storage system, and the diesel generator within the system; the start-up and shutdown costs of the micro-storage system are mainly determined by the operating status of the batteries, supercapacitors, and diesel generators in the system; the cost of purchasing electricity and the revenue from selling electricity are mainly related to the electricity price on the grid. The operating cost of a micro-storage system can be expressed by the following formula: Among them, F1 is the micro-storage system in the highway tunnel during the scheduling cycle. Total operating costs within; F om The operating and maintenance cost of the micro-storage system during the scheduling cycle; F on The start-up and shutdown cost of energy storage devices during the scheduling cycle of a micro-storage system; F grid c represents the power interaction cost between the micro-storage system and the grid during the scheduling cycle; ba P represents the battery operating cost coefficient, expressed in yuan / kWh. BA,ch (t), P BA,dis (t) represents the battery charging and discharging power of the micro-storage system at scheduling time t, respectively, in kW; c sc P represents the operating cost coefficient for supercapacitors, expressed in yuan / kWh. SC,ch (t), P SC,dis (t) represents the charging and discharging power of the supercapacitor in the micro-storage system at scheduling time t, respectively, in kW; c wp c pv These are the operating cost coefficients for distributed wind power generation systems and distributed photovoltaic power generation systems, respectively, in yuan / kWh; P wp (t), P pv (t) represents the theoretical wind power of the wind power generation unit at dispatch time t and the theoretical photovoltaic output of the photovoltaic power generation unit at dispatch time t, respectively, in kW; c dg P represents the operating cost coefficient for diesel generators, expressed in yuan / kWh. DG (t) represents the power generation of the diesel generator unit in the micro-storage system at dispatch time t, in kW; c ba,on This represents the battery start-stop cost coefficient, expressed in yuan / kWh; k ba (t) represents the battery operating status of the micro-storage system at scheduling time t; 1 indicates the battery is operating, and 0 indicates otherwise. sc,on This refers to the start-up and shutdown cost coefficient for supercapacitors, expressed in yuan / kWh; k sc (t) represents the operating state of the supercapacitor in the micro-storage system at time t; it is 1 if the supercapacitor is operating, and 0 otherwise. dg,on The start-up and shutdown cost coefficient for diesel generators is expressed in yuan / kWh; kdg (t) represents the operating status of the diesel generator in the micro-storage system at scheduling time t; 1 if the diesel generator is running, 0 otherwise; b(t) represents the electricity price at scheduling time t when the system interacts with the grid, in yuan / kWh; P grid (t) represents the power exchanged between the system and the grid at time t, in kW. When the system purchases electricity from the grid, P grid (t) is a positive value, while when the system sells electricity to the grid, P grid (t) is negative.

[0063] F on The main consideration is the impact of charging and discharging power on the start-up and shutdown costs of energy storage devices. For the same start-up cycle, shallow charging and discharging versus deep charging and discharging have different effects on the lifespan of the energy storage system; that is, the start-up and shutdown costs differ between shallow charging and discharging and deep charging and discharging. The greater the charging and discharging energy, the greater the lifespan loss of the energy storage device, which translates to higher start-up and shutdown costs.

[0064] Understandably, when the total number of tunnels n in the micro-storage system is greater than 1, the F1 values ​​corresponding to each of the n tunnels can be added together.

[0065] For micro-storage systems in highway tunnels, grid-connected operation is supported by the power grid, thus ensuring good power supply reliability. However, once the grid fails, the system enters islanded operation mode, and its power supply reliability will face a severe test. To ensure that the system still maintains high power supply reliability during islanded operation, the expected power shortage during islanded operation is used as an evaluation index for system power supply reliability, with minimizing the expected power shortage as the optimization objective.

[0066] In some embodiments, method 10 further includes: obtaining the power grid state coefficient at each scheduling time within the scheduling period; when the power grid is faulty, the power grid state coefficient is 1, and when the power grid is operating normally, the power grid state coefficient is 0.

[0067] The expected power shortage EENS is expressed by Equation 8: Formula 8; Where γ(t) is the power grid state coefficient.

[0068] This step defines the formula for calculating the Expected Energy Shortage (EENS) in the multi-objective optimization model, namely Formula 8. Its core logic is to calculate the cumulative energy shortage caused by the system's own power generation capacity being unable to meet the full load demand under the assumption of grid failure (islanding operation).

[0069] The Expected Energy Not Supplied (EVS) is typically measured in kWh. It quantifies the total electrical energy that the system is expected to be unable to supply to the load under a specific dispatching scheme when facing a anticipated event (in this case, a full-cycle grid failure). γ(t) is the grid state coefficient, a 0-1 variable. It is defined as: γ(t) at dispatching time t assuming the grid is in a fault state. When assuming the power grid is normal, γ(t) can be determined through power outage warnings. For example, if a power outage is predicted to occur between 13:00 and 14:00 the following day, γ(t) can be set to 1 for the corresponding scheduling time and 0 for other scheduling times. If set to operate within the entire scheduling cycle... This means that the assessment focuses on the most severe scenario of islanded operation throughout the entire scheduling cycle. l (t) represents the total electrical load demand (kW) of the tunnel at scheduling time t, derived from the output of step 180. wp (t), P pv (t) represents the theoretical wind power output and theoretical photovoltaic power output (kW) at scheduling time t, derived from the outputs of steps 150 and 170. Note that this is the predicted available output, not the actual value. P BA,dis (t) and P SC,dis (t) represents the planned discharge power (kW) of the battery and supercapacitor at scheduling time t, respectively. This is the decision variable of the model, i.e., part of the scheduling scheme to be determined during the optimization process. P DG γ(t) represents the planned output power (kW) of the diesel generator at scheduling time t, and is also a decision variable in the model. The max function ensures that the calculation result is non-negative. If the total power supply capacity of the system is greater than the total load demand, the deficit is 0; otherwise, the positive value of the difference between the load and the power supply capacity is taken. If γ(t) is 0, the deficit is 0.

[0070] This step transforms the abstract concept of power supply reliability into a calculable and optimizable specific mathematical indicator (EENS). Formula 8 sets... This parameter forces the optimization process to consider the extreme case of a complete grid outage. The optimization objective is to minimize the Energy ENS (Electricity Expansion and Reduction), driving the model to find a scheduling scheme that maximizes tunnel power supply even under the most unfavorable conditions. Minimizing the ENS guides the model to instruct the energy storage system to charge more frequently when electricity prices are low or renewable energy sources are abundant, reserving energy for potential islanding events. Simultaneously, it also affects the start-stop schedule of diesel generators, ensuring sufficient available capacity at critical moments. By treating the ENS under islanding scenarios as an independent optimization objective, reliability is elevated to a level of importance equal to economy and environmental friendliness, meeting the operational requirements of tunnel infrastructure.

[0071] For micro-storage systems in highway tunnels, lower carbon emissions mean less electricity the system obtains from diesel generators and the grid, and more electricity is generated by using low-carbon photovoltaic and wind power systems. Therefore, reducing carbon emissions helps improve the system's environmental benefits and reduce carbon pollution. Thus, minimizing the system's carbon emissions is one of the optimization goals for micro-storage systems in highway tunnels.

[0072] For example, the total carbon emissions of the system can be expressed by the following formula: Where F2 represents the total carbon emissions of the micro-storage system in highway tunnels, and C grid This represents the equivalent carbon emissions from electricity purchased from the grid during the system's scheduling cycle, expressed in kg CO2eq; C dg For the diesel generators in the system during the scheduling cycle Carbon emissions within the region, expressed in kg CO2eq;e grid The carbon emission factor of the power grid is expressed in kg CO2eq / kWh; e dg This represents the carbon emission factor during the operation of a diesel generator, expressed in kg CO2eq / kWh.

[0073] Constraints are equations or inequalities that must be satisfied during the optimization process, representing the physical laws and operational limitations of the system. For example, constraints can be expressed by the following formula. It is understood that constraints must be satisfied for any given value t.

[0074] (1) Power balance constraint When a micro-storage system is running in a highway tunnel, the system must satisfy the power balance constraint at any given time. The power balance constraint can be expressed by the following formula: The definitions of the parameters are the same as those in the previous text and will not be repeated here. P dep (t) represents the amount of solar power and air volume abandoned by the system at scheduling time t, in kW.

[0075] (2) Constraints on the purchase and sale of electricity with the power grid When a micro-storage system is operating in a highway tunnel, it is necessary to meet the constraints on the power exchange with the power grid, ensuring that the power exchange is within its upper and lower limits, as shown in the following formula: Among them, P grid (t) represents the interaction power between the system and the grid at scheduling time t. This interaction power is positive when the system purchases electricity from the grid, and negative when the system purchases electricity from the grid. The unit is kW; Pgrid,max This represents the upper limit of the power exchanged between the system and the power grid, which is a preset fixed value, and the unit is kW.

[0076] (3) Output constraints of internal equipment in micro-storage system During the operation of the micro-storage system in a highway tunnel, each device within the system needs to meet its own output constraints. The processing constraints of each device at scheduling time t are shown in the following formula: In the formula: P wp,max P wp,min The upper and lower limits of the output of the wind power generation unit, in kW; P pv,max P pv,min These represent the upper and lower limits of the output power of the photovoltaic power generation unit, in kW; P represents the power output of the diesel generator in the system at time t, in kW. DG,max P DG,min These represent the upper and lower limits of the diesel generator's output power, in kW; P BA,max P BA,min These represent the upper and lower limits of the battery's charging and discharging power, in kW; P SC,max P SC,min These represent the upper and lower limits of the charging and discharging power of the supercapacitor, in kW. SOC BA (t), SOC SC (t) represent the state of charge (SOC) of the battery and supercapacitor at scheduling time t, respectively, which can be determined by the charging and discharging power from the start of the prediction cycle to the current scheduling time; BA ,max、SOC BA ,min represent the maximum and minimum state of charge (SOC) values ​​of the battery, respectively; SC ,max、SOC SC ,min represent the maximum and minimum state of charge (SOC) values ​​of the battery, respectively.

[0077] Step 200: Solve the model to obtain the final operation scheduling scheme.

[0078] This step is the final solution and decision-making process for the multi-objective operation optimization model of the highway tunnel micro-storage system established in step 190. The process aims to solve the model and, from numerous possible solutions, select a single, explicit, and executable day-ahead scheduling scheme that achieves the optimal balance among three conflicting objectives: total system operating cost, expected power shortage, and total system carbon emissions, while satisfying all physical and operational constraints. Step 200 typically employs a two-stage combined strategy of optimization solution and multi-criteria decision-making. For example, step 200 includes: Step 2001: The NSGA-II algorithm is used to solve the operation optimization model to obtain the Pareto optimal solution set of the trade-off relationship between the total system operating cost, the expected power shortage, and the total system carbon emissions.

[0079] NSGA-II is an advanced multi-objective evolutionary algorithm. It simulates the biological evolutionary process (selection, crossover, mutation) and searches the decision variable space through iterative optimization. The Pareto Optimal Set is the set of solutions obtained when the algorithm terminates. Any solution in this set satisfies the following condition: it cannot improve any other objective function value without worsening at least one other objective function value. The images of these solutions in the objective function space constitute the Pareto Front. At this stage, the role of the NSGA-II algorithm is to efficiently and globally explore the complex, possibly non-convex, solution space, automatically discovering and generating a series of non-dominated solutions. Each solution represents a complete scheduling plan, corresponding to a specific trade-off between the three objectives (e.g., plan A has the lowest cost but slightly lower reliability, plan B has the highest reliability but the highest carbon emissions, and plan C is relatively balanced among the three). The importance of the three objectives (economy, reliability, and environmental protection) is difficult to quantify in advance. NSGA-II does not require manual setting of objective weights and can objectively reveal all possible trade-offs, providing a complete selection space for subsequent scientific decision-making.

[0080] The input to this step includes the complete multi-objective optimization model established in step 190, including the mathematical expressions of the three objective functions, constraints, and the domains of the decision variables. The output includes a set of dozens or even hundreds of Pareto optimal solutions. Each solution is a data structure containing the sequence of values ​​for all decision variables within the scheduling period, as well as the corresponding values ​​of the three objective functions (total cost, EENS, and total carbon emissions).

[0081] Step 2002: The VIKOR multi-criteria decision method is used to select the optimal compromise solution from the Pareto optimal solution set as the final operation scheduling scheme.

[0082] The VIKOR method is a multi-attribute decision-making tool based on the ideal point method. It ranks and selects solutions by measuring how close each candidate solution is to the ideal and negative ideal solutions. Its function is to systematically evaluate and compare a large number of excellent solutions generated in the first stage, automatically selecting the optimal compromise solution that best balances overall performance and individual performance using a set of objective criteria.

[0083] The optimal compromise solution may not be absolutely optimal for any one objective, but it is the closest to the ideal solution and achieves the best balance among conflicting objectives, and is therefore the most likely to be accepted by decision-makers.

[0084] In step 2001, NSGA-II outputs a solution library, while actual execution requires a specific instruction set. In step 2002, VIKOR provides a rational tool to automatically and objectively select the final solution from the solution library. The input to this stage is the Pareto optimal solution set output by NSGA-II in the first stage, along with the specific values ​​of the three objective functions corresponding to each solution. The output is a single optimal compromise solution. This solution is the final runtime scheduling scheme.

[0085] By employing a combined strategy of NSGA-II and VIKOR, step 200 transforms the complex multi-objective optimization problem into an operable and implementable scheduling instruction generation process. This ensures that the final solution is not only theoretically optimized but also engineering-wise sound. This scheme specifies in detail the amount of electricity the diesel generator should generate, the amount of charging / discharging the battery and supercapacitor should do, the amount of electricity purchased or sold from the grid, and whether some renewable energy sources need to be abandoned at each scheduling moment within the future scheduling cycle. Guided by this scheme, the micro-storage system in highway tunnels can achieve safe, economical, low-carbon, and reliable optimized operation.

[0086] This application's embodiments introduce and dynamically calculate the tunnel shading loss coefficient, quantifying the time-varying shading effects of structures such as mountains and tunnel entrances into the photovoltaic power output prediction model. This fundamentally corrects the systematic overestimation bias of traditional predictions under complex tunnel terrain, thereby significantly improving the accuracy of photovoltaic power output prediction. Based on this accurate prediction data, the constructed and solved optimized scheduling model can generate operating schemes highly matched to actual resource conditions. This makes scheduling commands such as energy storage charging and discharging, diesel engine start-stop, and grid interaction more reasonable in timing and more accurate in power balance, ultimately achieving a synergistic improvement in system operating economy, power supply reliability, and the practicality of the scheduling scheme.

[0087] In some embodiments, step 160 includes: Step 1601: Based on the height of the mountain above the tunnel, the size and orientation of the photovoltaic array, the azimuth angle of the tunnel, the solar altitude angle and solar azimuth angle at scheduling time t, calculate the ratio of the area of ​​the photovoltaic array covered by the mountain shadow to the total area of ​​the photovoltaic array at scheduling time t using the principle of geometric projection.

[0088] This step involves the quantitative calculation of the mountain's shadow occupancy ratio. Its core principle is to rely on geometric projection to integrate key parameters affecting the mountain's shadow for precise measurement. Specifically, the input parameters include the actual height of the mountain behind the tunnel, the specifications and location of the photovoltaic array, the tunnel's azimuth, and the solar altitude and azimuth angles corresponding to the scheduling time t. By constructing a geometric model, the projected area of ​​the mountain onto the photovoltaic array plane under sunlight is simulated, ultimately calculating the ratio of the area of ​​the photovoltaic array covered by the mountain's shadow at that moment to the total area of ​​the array.

[0089] Step 1602: Determine the tunnel shading loss coefficient at scheduling time t based on the ratio of the area of ​​the photovoltaic array covered by shadow to the total area of ​​the photovoltaic array, so as to obtain the tunnel shading loss coefficient at each scheduling time within the scheduling cycle.

[0090] This step is the derivation of the shading loss coefficient. Its core logic is based on the shadow area ratio obtained in step 1601, combined with the power generation characteristics of photovoltaic modules, to establish the correlation between the area shading ratio and the power generation efficiency loss, and then determine the tunnel shading loss coefficient corresponding to scheduling time t. By repeating the above calculation for each scheduling time within the scheduling cycle, the sequence of shading loss coefficients for each time point in the entire cycle can be obtained.

[0091] Specifically, step 1601 includes: Calculate the ratio of the area of ​​the photovoltaic array covered by the mountain shadow at the scheduling time t to the total area of ​​the photovoltaic array according to Formula 1; Formula 1.

[0092] Among them, A blo (t) represents the area of ​​the photovoltaic array covered by shadow at scheduling time t, A tol (t) Total area of ​​the photovoltaic array, L pv L is the length of the photovoltaic array along the tunnel entrance, and d is the horizontal distance from the foot of the mountain to the near edge of the photovoltaic array; shade Let t be the length of the shadow of the mountain from its foot to the end of the shadow that is far from the foot of the mountain; ; Let be the solar azimuth angle at time t. The azimuth is the clockwise angle between the tunnel exit direction and due south. h(t) is the solar altitude angle; H is the height of the mountain.

[0093] Figure 2 This is a schematic diagram illustrating the calculation principle of the tunnel shading loss coefficient in an embodiment of this application. Figure 3This is a schematic diagram showing the azimuth, solar azimuth angle, and tunnel azimuth angle of the photovoltaic array in an embodiment of this application. Figure 2 As shown, the horizontal distance from the foot of the mountain to the near edge of the photovoltaic array is d, and the length of the photovoltaic array along the tunnel entrance is L. pv When the length of the mountain shadow is less than or equal to d, the mountain shadow does not cover the photovoltaic array, and the area of ​​the photovoltaic array blocked by the mountain shadow is 0; when the length of the mountain shadow is greater than d+L... pv When the length of the mountain shadow is greater than d and less than or equal to d+Lpv, the photovoltaic array is partially blocked by the mountain shadow. The length of the blocked part of the photovoltaic array can be estimated by subtracting the length of the mountain shadow from the length of the blocked part, and then the ratio of the length of the blocked part to the length of the photovoltaic array can be used as the proportion of being blocked.

[0094] Understandably, since the photovoltaic arrays at the tunnel entrance are mostly strip-shaped, and their width is much smaller than the width of the mountain, the calculations assume that the shadows will completely cover the width of the photovoltaic arrays. (The sentence is incomplete and ends abruptly.) The width of the photovoltaic array in the middle is approximately Wpv.

[0095] Understandably, the calculations in this paper are based on the mountain above the tunnel being the main mountain, with the photovoltaic array laid out horizontally and centrally along the tunnel entrance. To simplify the model, the height of the photovoltaic panels is approximated as 0. The same method can be used to calculate photovoltaic tunnels and photovoltaic panels on slopes.

[0096] like Figure 2 As shown, the photovoltaic array can only be shaded when the sun is on the opposite side of the vertical section between the mountain and the tunnel. Therefore, it can be considered that when... Only then is it necessary to calculate the area of ​​the photovoltaic array that is blocked by the mountain.

[0097] Mountain shading is the most common and potentially most widespread source of shading in tunnel scenarios. Accurately calculating its shadow area is fundamental to assessing the overall impact of shading and is a core element in improving the accuracy of photovoltaic (PV) forecasts. Ignoring this component will lead to a significant overestimation of PV output during periods of low solar altitude, such as sunrise, sunset, or winter. This step utilizes the principle of geometric projection to calculate the shadow area cast onto the PV array plane by the mountain (mountain) near the tunnel at a specific solar position.

[0098] Step 1602 includes: Step 1602A: Calculate the occlusion ratio K at scheduling time t according to Formula 2. shade (t); Formula 2; Step 1602B: Calculate the tunnel shading loss coefficient k(t) at scheduling time t according to Formula 3; Formula 3.

[0099] Wherein, η is the ratio of the photovoltaic output when the array is completely shaded by the mountain to the photovoltaic output when it is not shaded under the same solar irradiance. This value can be obtained through field experiments and stored as an inherent parameter of the system. This parameter can be updated periodically (e.g., monthly).

[0100] The inventors of this application discovered that the shading of a mountain on a photovoltaic array only affects the direct component of sunlight, and does not affect the scattered and reflected components. By using η, the output effect of the photovoltaic array caused by the shading of the mountain can be limited to the effect on the direct sunlight, thus making the calculation of photovoltaic output more accurate.

[0101] In some embodiments, the weather forecast data also includes ambient temperature, and the system's inherent parameters also include the nominal cell operating temperature of the photovoltaic module. The method further includes: Step 210: Calculate the actual temperature of the photovoltaic array at scheduling time t within the scheduling cycle according to Formula 4; Formula 4.

[0102] Among them, T c (t) represents the actual temperature (°C) of the photovoltaic array at scheduling time t, indicating the true temperature of the photovoltaic cells under operating conditions, which directly affects its power generation efficiency. e G(t) represents the ambient temperature (°C) at scheduling time t, and G(t) represents the solar irradiance (W / m²) at scheduling time t. NOCT is the nominal operating cell temperature (°C) of the photovoltaic module, a key material and structural parameter that reflects the temperature rise characteristics (the increase in temperature relative to the ambient temperature) of this type of photovoltaic module under standard operating conditions. NOCT is the stable operating temperature reached by the photovoltaic module under specific standard environments (irradiance 800W / m², ambient temperature 20°C, wind speed 1m / s, etc.), provided by the manufacturer, and is usually between 40°C and 48°C.

[0103] The power generation efficiency of photovoltaic (PV) cells is negatively correlated with temperature; higher temperatures typically result in lower output voltage and maximum power point, leading to decreased power output. Therefore, accurately predicting module temperature is a necessary prerequisite for accurately calculating its actual power generation capacity (i.e., theoretical output). Ignoring the temperature effect, especially under high-temperature and high-irradiance conditions, will lead to an overestimation of PV output. This step establishes a mapping model from environmental meteorological conditions (ambient temperature, irradiance) to the core operating temperature of the PV module (cell temperature). This model captures the physical process of the module heating up due to the absorption of solar radiation.

[0104] Step 220: The product of the difference between the actual temperature of the photovoltaic array at scheduling time t and the temperature of the photovoltaic array under standard test conditions and the preset temperature coefficient is used as the temperature correction coefficient at scheduling time t; the specific calculation process is shown in the following formula.

[0105] in, The actual temperature of the component calculated in step 210; The component temperature is under standard test conditions, typically 25°C. The preset temperature power factor (unit: % / ℃) is provided by the photovoltaic module manufacturer and is usually a negative value (e.g., -0.4% / ℃), representing the percentage decrease in power for every 1℃ increase in temperature.

[0106] To derive the theoretical photovoltaic output reflecting actual operating conditions from the ideal photovoltaic output calculated based on STC conditions, temperature effects must be compensated for. Temperature correction factors provide a quantitative tool for this compensation. Combining them with shading loss factors allows for a dual, refined correction of photovoltaic output.

[0107] Accordingly, step 170 includes: correcting the ideal photovoltaic output at scheduling time t using the tunnel shading loss coefficient and the temperature correction coefficient at scheduling time t, to obtain the theoretical photovoltaic output at scheduling time t.

[0108] For example, the ideal photovoltaic output at scheduling time t is compared with the tunnel shading loss coefficient k(t) at scheduling time t and the temperature correction coefficient k(1+schedule time t). temp Multiplying (t) by the power of the photovoltaic system at time t, we obtain the theoretical photovoltaic output P. pv (t).

[0109] Photovoltaic output in tunnel scenarios is affected by both the reduction in available light due to shading from mountains and the decrease in conversion rate due to internal electrothermal characteristics (higher temperature leads to lower efficiency). Considering only shading is incomplete. This step couples these two factors, achieving a more comprehensive model of the impact mechanism on output. This dual correction significantly improves the adaptability of the prediction model in different seasons (large temperature differences), different weather conditions (variable combinations of irradiance and temperature), and the special microenvironment of tunnels. This results in a higher level of agreement between the predicted curve and actual operating data, fundamentally enhancing the accuracy and reliability of subsequent optimization and scheduling schemes.

[0110] In this embodiment, two core influencing factors, spatial physical shading and electrothermal characteristic attenuation, are integrated to make comprehensive and refined corrections to the ideal photovoltaic output, thereby generating a highly reliable theoretical photovoltaic output prediction sequence for system optimization and scheduling.

[0111] In some embodiments, the system's inherent parameters also include the fixed load of the monitoring and safety system, and the calculation of the total tunnel electrical load at each scheduling time within the scheduling cycle, including: Step 1801: Calculate the lighting load at each scheduling time within the scheduling cycle; the lighting load includes a fixed portion and a variable portion that is positively correlated with the traffic flow forecast data.

[0112] The fixed portion maintains the basic lighting power required for the minimum safe operation of the tunnel. This power remains essentially constant throughout the scheduling cycle and corresponds to the power consumed by the tunnel's 24-hour on-call lights or the lowest brightness level lighting, unaffected by traffic flow. This is the electrical load required to ensure basic traffic safety in the tunnel. The variable portion is the additional lighting power dynamically adjusted according to real-time traffic flow. This power is positively correlated with predicted traffic flow and aims to provide drivers with brightness appropriate to traffic density and vehicle speed to ensure driving safety and comfort. This corresponds to the power consumed by dimmable lamps. Its magnitude is automatically adjusted according to management strategies and is directly related to traffic flow. The greater the traffic flow and the faster the vehicle speed, the higher the drivers' lighting requirements, and the greater the load on this portion.

[0113] The calculation of the variable part is usually based on a normalized traffic flow function. (Values ​​range from 0 to 1) and the maximum adjustable power P of the lighting system. light,adj Therefore, the lighting load at time t It can be represented as: P light (t) = P light,base + β l f light (Q(t)) P light,adj Among them, P light,base The reference lighting power is expressed in kW. β l This represents the actual percentage of the adjustable portion; some light fixtures are not fully adjustable, 0 < β l ≤1; f light (Q(t)) is the traffic flow normalization function, taking values ​​in [0, 1]; P light,adj This represents the total rated power of the adjustable section, expressed in kW.

[0114] The traffic flow normalization function is expressed by the following formula: Where Q(t) represents the traffic flow at time t, in units of... ; and Q max Q min These represent the system's maximum and minimum traffic volumes, respectively.

[0115] Understandably, the reference lighting power, the actual proportion of the adjustable portion, the maximum traffic flow, and the minimum traffic flow are inherent parameters of the system.

[0116] The lighting load includes a fixed component and a variable component that is positively correlated with traffic flow forecast data, enabling dynamic prediction of the power consumption of the tunnel lighting system. By distinguishing between the fixed and variable components, the model can reflect both the basic energy consumption that the tunnel must maintain and accurately capture the characteristics of load fluctuations with traffic conditions.

[0117] Step 1802: Calculate the ventilation load at each scheduling time within the scheduling cycle; the ventilation load includes a fixed component and a variable component that is positively correlated with the traffic flow forecast data.

[0118] The calculation of ventilation load is also divided into two parts. The fixed part is the constant base load power of the ventilation fans required to maintain the minimum air quality in the tunnel. The variable part is the additional ventilation power required to dilute and remove pollutants (such as CO and NOx) and heat emitted by vehicles. This part of the power is closely positively correlated with traffic flow, because the rate of pollutant and heat generation directly depends on the number of vehicles passing through the tunnel.

[0119] Ventilation load can be expressed as: P vent (t) represents the power of the fan at time t, in kW; P fan,base P represents the ventilator's permanent base load power, measured in kW. fan (t) represents the variable power of the ventilator, which is affected by traffic flow and air pollutant concentration, and is expressed in kW.

[0120] Ventilation is typically the largest electrical load within a tunnel, and it exhibits the strongest fluctuations. The response of ventilation load to traffic flow is a core characteristic of tunnel energy consumption. Scheduling based on average or maximum load would result in significant energy waste or operational risks. Modeling ventilation load as a function of traffic flow allows for accurate prediction of ventilation load, which is crucial for optimizing the allocation of sufficient and economical power generation and energy storage resources in scheduling.

[0121] In some embodiments, the system inherent parameters further include fan efficiency; in step 1802, calculating the ventilation load at each scheduling moment within the scheduling cycle includes: According to Formula 5, calculate the variable part of the ventilation load at scheduling time t within the scheduling cycle; Formula 5; in, (t) represents the pressure difference at time t during operation of the fan, measured in Pa. It refers to the total pressure difference between the fan inlet and outlet, and is the power source required to drive airflow over resistance within the tunnel. It is determined by frictional resistance along the tunnel and the piston-wind effect in traffic. Q v (t) represents the required ventilation volume of the fan at scheduling time t, expressed in m³ / s. It refers to the volumetric airflow rate that needs to be exhausted from the tunnel at scheduling time t to control the pollutant concentration and temperature within safe and comfortable standards. It primarily depends on the total amount of pollutants and heat emitted by vehicles, and therefore is related to traffic flow. Directly related. Fan efficiency refers to the efficiency with which a fan converts input electrical energy into effective air power (pressure difference × flow rate). It takes into account mechanical, electrical, and fluid losses and is an inherent parameter of the system, and is a dimensionless number.

[0122] Formula 5, based on the fundamental principles of fluid mechanics (power = pressure × flow rate / efficiency), defines the physical requirements for ventilation (flow rate). ) and system resistance (pressure difference) The electricity demand is directly converted into the power demand of the driving equipment (wind turbine). This power represents the variable portion of the ventilation load. Modeling directly using physical formulas more fundamentally reflects the intrinsic relationship between ventilation system energy consumption, traffic flow, and tunnel geometry parameters, achieving higher prediction accuracy than simple empirical linear relationships. This allows the optimized scheduling model to more accurately assess the enormous ventilation energy consumption during peak traffic periods, thereby enabling more rational power generation and energy storage scheduling arrangements.

[0123] In some embodiments, the system inherent parameters also include air density, friction coefficient, tunnel length, tunnel equivalent diameter, tunnel cross-sectional area, and piston wind coefficient, which are calculated in step 1802 according to formula 6. ; Formula 6; in, For air density, λ, L, D h These represent the friction coefficient along the tunnel, tunnel length, and equivalent tunnel diameter, respectively; A is the tunnel cross-sectional area, and k... tr Let Q(t) be the piston wind coefficient, and Q(t) be the traffic flow prediction data at scheduling time t.

[0124] Formula 6, first term on the right side of the equal sign Frictional resistance term along the path, where The friction coefficient is the coefficient along the tunnel wall, which characterizes the resistance of the tunnel wall roughness to the airflow and is determined by the tunnel wall material. The length of the tunnel is in meters (m). The equivalent diameter of the tunnel (m) is a comprehensive geometric parameter that measures the influence of the tunnel cross-sectional shape on flow. This refers to the air density (kg / m³). The cross-sectional area of ​​the tunnel is (m²). Proportional to the square of the average wind speed inside the tunnel, it represents the dynamic pressure loss caused by the airflow velocity.

[0125] Formula 6, second term on the right side of the equation This is the piston wind effect term. The piston wind coefficient (Pa / veh) is an empirical coefficient used to quantify the piston effect that a vehicle exerts on the air inside a tunnel. As a vehicle moves forward, it pushes the air in front of it, creating a negative pressure zone behind the vehicle, thus generating additional ventilation force or drag. The traffic flow prediction data (veh / h) is for the scheduling time t.

[0126] In this step, the total pressure difference is decomposed into two main parts. The first part is static fluid resistance, which is the constant-direction resistance generated by the friction between air and the tunnel walls as air flows through the tunnel. This resistance is proportional to the square of the ventilation volume and constitutes the main part of energy consumption. The second part is dynamic traffic impact, which is the additional effect of vehicle movement on airflow. This part is related to traffic volume. The linear relationship reflects the direct dynamic coupling between load and traffic flow.

[0127] Compared to using a simple overall coefficient, decomposing the pressure difference into friction and piston-wind terms provides a more accurate description of the actual resistance characteristics of different tunnels (of varying lengths and roughness) under different ventilation modes and traffic conditions. The introduction of the piston-wind term is crucial. It reflects the nonlinearity and even directionality of the impact of traffic flow on ventilation load: under specific conditions (such as one-way peak traffic flow), a large number of vehicles may help ventilation, thereby reducing fan energy consumption; while in opposite traffic flow or low-speed congestion, it may increase turbulence and resistance. Ignoring this term will lead to systematic biases in load forecasting.

[0128] For example, in step 1802, the required ventilation volume of the fan at scheduling time t is calculated as follows: wherein the required ventilation volume of the fan at scheduling time t is the larger of the ventilation volume required to maintain the air pollutant concentration in the tunnel within the pollutant limit value and the ventilation volume required to maintain the temperature rise in the tunnel within the safety limit value at scheduling time t.

[0129] Specifically, the required ventilation volume of the fan at scheduling time t is calculated according to formula 7; Formula 7; Among them, e poll The pollutant emission rate per vehicle; C maxC0 represents the maximum permissible concentration of pollutants in the tunnel, while C0 represents the concentration of pollutants in the air outside the tunnel during the scheduling cycle, which can be considered a fixed value throughout the entire scheduling cycle; h represents the thermal power of each vehicle. For air density, c p ΔT is the specific heat capacity of air. max This represents the maximum acceptable temperature rise for the tunnel.

[0130] The system's inherent parameters also include the pollutant mass emission rate and thermal power of each vehicle, the maximum allowable concentration of pollutants in the tunnel, the specific heat capacity of the air, the air density, and the maximum acceptable temperature rise in the tunnel; meteorological forecast data includes the pollutant concentration of the air outside the tunnel itself during the scheduling cycle.

[0131] This application focuses on the key parameter in calculation formula 5, namely the ventilation volume required by the fan at scheduling time t. This parameter directly determines the air exchange intensity necessary to maintain a safe and healthy environment within the tunnel. This application employs a constraint-based engineering design method oriented towards the most unfavorable operating conditions to determine this parameter. This ensures that the ventilation system is capable of addressing the two core environmental challenges of pollutant diffusion and heat accumulation.

[0132] The required ventilation volume of the fan It is not an arbitrary value, but is determined by two rigid constraints controlling the tunnel environment: pollutant concentration limits and temperature rise limits. During the calculation, the required ventilation volume is first calculated based on these two constraints, and then the larger of the two values ​​is taken as the final value. This is specifically achieved through Formula 7.

[0133] in This refers to the ventilation volume based on pollutant constraints, which is the minimum ventilation volume required to dilute pollutants emitted by vehicles (such as carbon monoxide (CO), nitrogen oxides (NOx), particulate matter, etc.) to below a safe concentration in order to ensure air quality inside the tunnel.

[0134] The traffic flow forecast data (vehicles / hour) at scheduling time t represents the source intensity of pollutants. The pollutant mass emission rate (kg / veh) represents the average mass of specific pollutants emitted by each vehicle while passing through a tunnel. It depends on the vehicle type, fuel, and driving conditions and can be calculated using the emission rate of a typical passenger car. This represents the maximum permissible concentration (kg / m³) of pollutants in the tunnel. This represents the background concentration (kg / m³) of pollutants in the air during the dispatch cycle.

[0135] This refers to the minimum ventilation volume required to control the temperature rise inside the tunnel caused by vehicle heat dissipation (engine, brakes, tires, etc.) and prevent excessively high temperatures from affecting equipment operation, road conditions, or the comfort of drivers and passengers.

[0136] The thermal power per vehicle (kW / veh) represents the average amount of heat emitted by each vehicle to the environment while it is running in the tunnel. It can be calculated using the thermal power of a typical passenger car. This refers to the air density (kg / m³). is the specific heat capacity of air (J / (kg·K)), which represents the amount of heat required to raise the temperature of a unit mass of air by 1 degree Celsius. This indicates the maximum acceptable temperature rise (°C) inside the tunnel, which is the maximum permissible increase in air temperature inside the tunnel relative to the outside air temperature at the entrance.

[0137] In this scheme, by taking the maximum value, it identifies whether pollutant control or temperature control becomes the decisive factor in ventilation demand at a specific time t. This ensures that the ventilation volume meets all safety requirements at all times. The formula clearly shows that the required ventilation volume is related to the traffic flow. It is directly proportional. The greater the traffic volume, the more pollutants and heat are generated, and the greater the amount of ventilation air required for dilution or cooling. The calculation using this method... This accurately reflects the ventilation power required to meet dynamic environmental standards, providing a basis for subsequent accurate calculations of ventilation load. It provides core inputs, avoiding the overestimation of energy consumption during off-peak periods or underestimation of energy consumption during peak periods that can result from using a fixed ventilation volume.

[0138] Step 1803: Add the lighting load, ventilation load, and monitoring and safety system fixed load at each scheduling time within the scheduling cycle to obtain the total tunnel electrical load at each scheduling time within the scheduling cycle.

[0139] Perform this summation operation for each scheduling time t within the scheduling period to generate a predicted time series of total load.

[0140] Among them, the fixed load of the monitoring and security system is the inherent load of the system. It is the constant power consumption of the security systems in the tunnel, such as monitoring (e.g., cameras), communication, fire protection, and alarm, under normal operating conditions (without special emergency modes such as fire or accident). This load is considered a constant value within the scheduling cycle.

[0141] Unlike conventional static load estimation, this embodiment fully considers the typical characteristics of highway tunnel load changes with traffic flow, decomposes the total load into dynamic components (variable parts of lighting and ventilation) that are strongly correlated with traffic flow and relatively fixed basic components, thereby establishing an accurate load prediction model and providing accurate power demand input for subsequent optimized scheduling.

[0142] Principles and steps not explicitly described in this invention are all obtainable by those skilled in the art through conventional technical means, and therefore will not be elaborated upon. Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for optimizing the operation and scheduling of a micro-storage system in a highway tunnel, characterized in that, The tunnel micro-storage system includes a photovoltaic power generation unit, a wind power generation unit, a diesel power generation unit, a battery energy storage unit, a supercapacitor energy storage unit, and a grid interaction interface; the photovoltaic power generation unit includes a photovoltaic array; the photovoltaic array is constructed along the tunnel entrance direction; the method includes: The inherent parameters of the highway tunnel micro-storage system during the scheduling cycle are obtained; the inherent parameters of the system include the height of the mountain above the tunnel, the size and orientation of the photovoltaic array, the azimuth angle of the tunnel, the power characteristic curve of the wind turbine, and the power parameters of the photovoltaic array. Obtain the solar altitude angle and solar azimuth angle at each scheduling moment within the scheduling cycle; Obtain meteorological forecast data for each scheduling moment within the scheduling cycle of the tunnel location; the meteorological forecast data includes solar irradiance, wind direction, and wind speed; Obtain traffic flow prediction data for each scheduling time within the scheduling cycle of the tunnel location; Based on the wind direction, wind speed, and the power characteristic curve of the wind turbine, the theoretical wind power output of the wind power generation unit at each scheduling moment within the scheduling cycle is calculated; based on the solar irradiance and the power parameters of the photovoltaic array, the ideal photovoltaic output of the photovoltaic power generation unit at each scheduling moment within the scheduling cycle is calculated. Calculate the tunnel shading loss coefficient at each scheduling time within the scheduling cycle; the tunnel shading loss coefficient is used to characterize the degree to which the photovoltaic output of the photovoltaic array is reduced due to the shadow of the mountain. For scheduling time t within the scheduling cycle, the ideal photovoltaic output at scheduling time t is corrected using the tunnel shading loss coefficient at scheduling time t to obtain the theoretical photovoltaic output at scheduling time t, thereby obtaining the theoretical photovoltaic output at each scheduling time within the scheduling cycle. Based on the traffic flow prediction data and the inherent parameters of the system, the total electrical load of the tunnel at each scheduling time within the scheduling cycle is calculated. Based on the theoretical photovoltaic output, theoretical wind power output, and total tunnel electrical load at each scheduling moment within the scheduling cycle, and with system power balance and equipment operation constraints as constraints, and with the optimization objectives of minimizing total system operating cost, minimizing expected power shortage, and minimizing total system carbon emissions, an operation optimization model for a highway tunnel micro-storage system is established. The decision variables of the model include the battery charging power, battery discharging power, supercapacitor charging power, supercapacitor discharging power, diesel generator output power, wind curtailment power, solar curtailment power, and the power exchanged with the grid through the grid interaction interface at each scheduling moment within the scheduling cycle. The model is solved to obtain the final operation scheduling scheme; The calculation of the tunnel shading loss coefficient at each scheduling moment within the scheduling cycle includes: For scheduling time t within the scheduling cycle, based on the height of the mountain above the tunnel, the size and orientation of the photovoltaic array, the azimuth angle of the tunnel, the solar altitude angle and solar azimuth angle at scheduling time t, the ratio of the area of ​​the photovoltaic array covered by the mountain shadow at scheduling time t to the total area of ​​the photovoltaic array is calculated using the principle of geometric projection. Based on the ratio of the area of ​​the photovoltaic array covered by shadow to the total area of ​​the photovoltaic array at the scheduling time t, the tunnel shading loss coefficient at scheduling time t is determined, so as to obtain the tunnel shading loss coefficient at each scheduling time within the scheduling cycle.

2. The method according to claim 1, wherein, For scheduling time t within the scheduling cycle, based on the height of the mountain above the tunnel, the size and orientation of the photovoltaic array, the azimuth angle of the tunnel, and the solar altitude and azimuth angle at scheduling time t, the ratio of the area of ​​the photovoltaic array covered by the mountain shadow at scheduling time t to the total area of ​​the photovoltaic array is calculated using the principle of geometric projection, including: Calculate the ratio of the area of ​​the photovoltaic array covered by the mountain shadow at the scheduling time t to the total area of ​​the photovoltaic array according to Formula 1; Official 1; Among them, A blo (t) represents the area of ​​the photovoltaic array covered by shadow at scheduling time t, A tol (t) Total area of ​​the photovoltaic array, L pv L is the length of the photovoltaic array along the tunnel entrance, and d is the horizontal distance from the foot of the mountain to the near edge of the photovoltaic array; shade Let t be the length of the shadow of the mountain from its foot to the end of the shadow that is far from the foot of the mountain; ； Let be the solar azimuth angle at time t. The azimuth is the clockwise angle between the tunnel exit direction and due south. h(t) is the solar altitude angle; H is the height of the mountain. The tunnel shading loss coefficient at scheduling time t is determined based on the ratio of the area of ​​the photovoltaic array covered by shadow to the total area of ​​the photovoltaic array, including: The occlusion ratio K at scheduling time t is calculated using Formula 2. shade (t); Official 2; The tunnel shading loss coefficient k at scheduling time t is calculated according to Formula 3. Official 3.

3. The method according to claim 1 or 2, wherein, The meteorological forecast data also includes ambient temperature, and the system's inherent parameters also include the nominal cell operating temperature of the photovoltaic modules; the method further includes: The actual temperature of the photovoltaic array at scheduling time t within the scheduling cycle is calculated according to Formula 4. Official 4; Among them, T c (t) represents the actual temperature of the photovoltaic array at scheduling time t; T e G(t) represents the ambient temperature at scheduling time t, G(t) represents the solar irradiance at scheduling time t, and NOCT represents the nominal cell operating temperature of the photovoltaic module. The product of the difference between the actual temperature of the photovoltaic array at scheduling time t and the temperature of the photovoltaic array under standard test conditions and the preset temperature coefficient is used as the temperature correction coefficient at scheduling time t. The ideal photovoltaic output at scheduling time t is corrected using the tunnel shading loss coefficient at scheduling time t, and the theoretical photovoltaic output at scheduling time t is obtained by correcting the ideal photovoltaic output at scheduling time t using the temperature correction coefficient at scheduling time t.

4. The method according to claim 1, wherein, The system's inherent parameters also include the fixed load of the monitoring and safety system; the calculation of the total tunnel electrical load at each scheduling moment within the scheduling cycle includes: Calculate the lighting load at each scheduling time within the scheduling cycle; the lighting load includes a fixed portion and a variable portion that is positively correlated with traffic flow forecast data; Calculate the ventilation load at each scheduling time within the scheduling cycle; the ventilation load includes a fixed component and a variable component that is positively correlated with the traffic flow forecast data; The total electrical load of the tunnel at each scheduling time within the scheduling cycle is obtained by adding the lighting load, ventilation load, and monitoring and safety system fixed load at each scheduling time within the scheduling cycle.

5. The method according to claim 4, wherein, The system's inherent parameters also include fan efficiency, and the calculation of the ventilation load at each scheduling moment within the scheduling cycle includes: According to Formula 5, calculate the variable part of the ventilation load at scheduling time t within the scheduling cycle; Official 5; in, (t) is the pressure difference during wind turbine operation at scheduling time t; Q v (t) represents the required ventilation volume of the fan at scheduling time t. For fan efficiency.

6. The method according to claim 5, wherein, The system's inherent parameters also include air density, friction coefficient, tunnel length, equivalent tunnel diameter, tunnel cross-sectional area, and piston wind coefficient; the calculation of the ventilation load at each scheduling moment within the scheduling cycle also includes: According to Formula 6, the calculation is as follows: p; Official 6; in, For air density, λ, L, D h These represent the friction coefficient along the tunnel, tunnel length, and equivalent tunnel diameter, respectively; A is the tunnel cross-sectional area, and k... tr Let Q(t) be the piston wind coefficient, and Q(t) be the traffic flow prediction data at scheduling time t.

7. The method according to claim 5, wherein, The calculation of the ventilation load at each scheduling time within the scheduling cycle also includes: calculating the ventilation volume required by the fan at scheduling time t; wherein the ventilation volume required by the fan at scheduling time t is the larger of the ventilation volume required to maintain the air pollutant concentration in the tunnel within the pollutant limit value and the ventilation volume required to maintain the temperature rise in the tunnel within the safety limit value at scheduling time t.

8. The method according to claim 7, wherein, The inherent parameters of the system also include the pollutant mass emission rate and thermal power of each vehicle, the maximum allowable concentration of pollutants in the tunnel, the specific heat capacity of the air, the air density, and the maximum acceptable temperature rise in the tunnel; the meteorological forecast data includes the pollutant concentration of the air outside the tunnel itself during the scheduling cycle; the ventilation volume required by the fan at the scheduling time t is calculated according to Formula 7; Official 7; Among them, e poll The pollutant emission rate per vehicle; C max C0 represents the maximum permissible concentration of pollutants in the tunnel, while h represents the concentration of pollutants in the air outside the tunnel during the scheduling cycle. For air density, c p ΔT is the specific heat capacity of air. max This represents the maximum acceptable temperature rise for the tunnel.

9. The method according to claim 7, wherein, The method further includes: Obtain the power grid state coefficient at each scheduling moment within the scheduling cycle; when the power grid is faulty, the power grid state coefficient is 1, and when the power grid is operating normally, the power grid state coefficient is 0. The minimum expected power shortage EENS is expressed by Equation 8: Official 8; Where γ(t) is the power grid state coefficient.

10. The method according to claim 1, wherein, Solving the model yields the final runtime scheduling scheme, which includes: The NSGA-II algorithm is used to solve the operation optimization model to obtain the Pareto optimal solution set of the trade-off relationship between the total system operating cost, the expected power shortage, and the total system carbon emissions; The optimal compromise solution is selected from the Pareto optimal solution set using the VIKOR multi-criteria decision method as the final operation scheduling scheme.

Images