Microgrid Control Method for Highway Service Areas Based on Distributed Photovoltaic Energy Storage

Abstract

This invention relates to the field of microgrid control and discloses a method for regulating a highway service area microgrid based on distributed photovoltaic energy storage. The method includes the following steps: First, traffic flow data is collected to construct a traffic-energy potential field model, mapping discrete vehicles to a continuous potential field to predict net power demand. Second, a kinetic-thermal entropy mapping is constructed using vehicle kinetic energy characteristics, and battery thermal management feedforward control is executed before load arrival. Simultaneously, variational mode decomposition parameters are adaptively adjusted according to the potential field gradient to decompose the power command into high-frequency and mid-to-low-frequency components. The high-frequency components are absorbed by supercapacitors, while the mid-to-low-frequency components are collaboratively distributed among battery packs based on a virtual complex impedance model constructed using health status and temperature. This invention achieves proactive defense against impact loads and refined lifespan extension of energy storage through traffic-energy cross-domain coupling, significantly improving the voltage rigidity and operational stability of the microgrid under extreme conditions.

Description

Technical Field

[0001] This invention relates to the field of microgrid control technology, specifically to a method for regulating a highway service area microgrid based on distributed photovoltaic energy storage. Background Technology

[0002] With the advancement of development goals and the explosive growth in the number of new energy vehicles, highway service areas are gradually transforming from simple rest areas into comprehensive micro-energy hubs integrating photovoltaic power generation, energy storage regulation, and high-power charging. To cope with the impact load brought by the high-power fast charging of electric vehicles, configuring hybrid energy storage systems has become the mainstream solution. However, in actual operation, highway service area microgrids face multiple challenges, including the high randomness of traffic flow, severe load fluctuations, and accelerated aging of energy storage equipment. Existing regulation technologies still have significant limitations in addressing these issues.

[0003] Existing service area microgrid energy management systems typically rely on passive adjustments based solely on real-time monitoring of local electrical quantities or statistical patterns of historical load data. This approach severs the inherent physical connection between the transportation network and the energy network, ignoring the fact that upstream high-speed vehicles are essentially arriving mobile high-energy loads. Due to the lack of proactive sensing of the potential energy of upstream traffic flow, the system often only initiates a response momentarily after the impact load actually connects to the bus. This lagging control strategy makes the DC bus voltage highly susceptible to sudden drops or fluctuations, severely impacting power quality and system stability.

[0004] At the power distribution and battery management level, existing technologies mostly employ filtering algorithms with fixed cutoff frequencies or rule-based logic control to allocate high-frequency and low-frequency power. However, highway charging loads exhibit extreme time-varying characteristics, making it difficult for fixed-parameter filters to simultaneously address the needs of mitigating instantaneous impacts and protecting battery life. This results in high-frequency peak power, which should be handled by supercapacitors, often penetrating to the battery terminals. More seriously, traditional battery thermal management systems generally employ hysteresis feedback control based on temperature thresholds, meaning that cooling measures are only activated after the battery temperature actually rises. This "remedial" approach to thermal management cannot offset the heat accumulation effect during the initial stages of high-rate discharge. Furthermore, when battery packs operate in parallel, existing control strategies rarely consider the inconsistencies between the health status and real-time temperature of individual battery modules. This leads to batteries in "poor condition" bearing equal or even excessive current, exacerbating the bottleneck effect within the battery pack and accelerating the performance degradation of the entire energy storage system. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method for regulating microgrids in highway service areas based on distributed photovoltaic energy storage. This method solves the problems of poor DC bus voltage stability and accelerated aging of energy storage batteries due to heat accumulation and uneven power distribution when existing highway service area microgrids face the impact of random high-power charging by electric vehicles. These problems arise from the lack of advanced perception of upstream traffic flow potential energy and the lag and crudeness of energy storage control strategies.

[0006] To achieve the above objectives, the present invention provides the following technical solution: The first aspect of the present invention provides a method for regulating a microgrid in a highway service area based on distributed photovoltaic energy storage. This method achieves advanced prediction and refined regulation of the grid load by introducing traffic flow potential energy field theory and multi-physics field coupling control.

[0007] Specifically, the method includes the following steps:

[0008] The system collects operational status data, environmental data, and traffic flow data of the service area microgrid and upstream road sections. The traffic flow data includes at least vehicle density, average vehicle speed, and electric vehicle penetration rate.

[0009] Based on the traffic flow data, a traffic-energy potential field model is established to calculate the potential field intensity within the future time window. Combined with the microgrid's operating status data and environmental data, a net power demand prediction curve for the microgrid is generated.

[0010] In this embodiment, the core of establishing the traffic-energy potential field model lies in mapping the microscopic vehicle driving behavior to the macroscopic energy impact pressure. Upstream electric vehicles are represented in the model as a flow of particles carrying energy demand. Using the definition of kinetic energy in physics, a weighted integral is performed on the kinetic energy corresponding to the vehicle density and average vehicle speed within a defined range upstream.

[0011] Preferably, time Traffic-energy potential field strength The calculation formula is expressed as follows:

[0012]

[0013] in, For electric vehicle penetration rate, For position Vehicle density at the location This is the normalization coefficient for the average vehicle mass. The average speed of the vehicle. and These are the starting and ending points of the monitoring range for the upstream road section, respectively. The traffic-energy potential field intensity... The numerical value intuitively represents the potential charging energy demand carried by the traffic flow that is about to arrive at the service area, and it is positively correlated with the upstream vehicle density, penetration rate and the square of speed.

[0014] Furthermore, when generating the net power demand forecast curve, a potential-power mapping function is obtained by fitting historical data through regression analysis. The field intensity is converted into a load forecast value, and the calculation formula is as follows:

[0015]

[0016] in, For net power requirements, For the service area's basic load, For the predicted power of photovoltaic power generation, To predict the lead time window.

[0017] One of the key innovations of this invention is the introduction of thermal management feedforward control. Based on the kinetic energy characteristics of the vehicle's average driving speed, a kinetic energy-thermal entropy mapping relationship is constructed to predict the heat generation risk of the energy storage battery, and pre-temperature control is performed on the thermal management system of the energy storage battery before the load corresponding to the net power demand prediction curve arrives.

[0018] Specifically, the square of the average speed of upstream vehicles is calculated as the average kinetic energy characteristic of the traffic flow. And utilize the kinetic energy-current-heat conversion coefficient Estimate the expected heat generation rate of the battery :

[0019]

[0020] in, This represents the current internal resistance of the battery. The physical meaning of this mapping relationship is that a higher upstream vehicle speed usually means that the vehicle has a lower remaining charge (SOC) and a higher charging rate requirement when it arrives at the service area. This results in the energy storage battery having to withstand high current discharge for a long time, which increases the risk of heat generation.

[0021] Based on this expected heat production rate, the target reference temperature of the thermal management system is calculated. :

[0022]

[0023] in, The optimal electrochemical reaction temperature for the battery (e.g., 25°C). For the specific heat capacity of the battery module, This refers to the thermal safety margin factor. Before the vehicle arrives, the control system pre-adjusts the battery temperature to below [a certain level]. of It utilizes the battery's own thermal capacity to absorb the impending thermal shock and prevent the temperature from exceeding the limit.

[0024] Another key innovation of this invention lies in gradient-based adaptive frequency domain decomposition. To accurately address power fluctuations with different characteristics, the gradient of the potential field intensity is calculated, and the parameters of the variational mode decomposition (VMD) algorithm are dynamically adjusted using this gradient to decompose the net power demand prediction curve into modal components of different frequencies.

[0025] Specifically, the gradient of the potential field change is obtained by calculating the derivative of the potential field intensity with respect to time. The quadratic penalty factor of the VMD algorithm is adaptively adjusted based on this gradient. :

[0026]

[0027] in, As the benchmark penalty factor, To adjust the gain. This formula shows that when the potential field changes gradient... When the factor increases (indicating a potential for a severe power surge), the second-order penalty factor... Decrease the value to widen the decomposition bandwidth and ensure the algorithm can capture high-frequency impulse components; conversely, increase the value. To filter out noise.

[0028] Subsequently, a cutoff frequency was set, and the decomposed modal components were divided into high-frequency power components. and low- and mid-frequency power section .Will Allocate resources to supercapacitors to leverage their high power density; It is allocated to the battery pack to take advantage of its high energy density.

[0029] The third innovation of this invention lies in the collaborative power distribution of the battery pack based on virtual complex impedance. To address the inconsistency problem in parallel battery packs, a virtual complex impedance model is constructed based on the state of health (SOH) of the energy storage batteries and real-time temperature.

[0030] For each set of parallel battery cells in a hybrid energy storage system Its virtual impedance value The construction model is as follows:

[0031]

[0032] in, The nominal impedance is... , These are the weighting coefficients. For real-time temperature, This is the temperature deviation dead zone threshold. This model causes battery cells with low SOH or large temperature deviations from the optimal range to exhibit high impedance characteristics.

[0033] In the allocation of low-frequency power section At that time, the principle of current division in parallel circuits is utilized:

[0034]

[0035] This enables adaptive dynamic power balancing, allowing batteries in poor condition to automatically take on less power, thereby extending the overall lifespan of the battery pack.

[0036] In addition, the method also includes a closed-loop feedback correction mechanism to monitor the bus voltage deviation in real time and control the tie line switching power when the deviation exceeds the threshold, while using actual arrival data to correct the parameters in the potential energy-power mapping function online.

[0037] A second aspect of the present invention provides a microgrid control system for highway service areas based on distributed photovoltaic energy storage, the system comprising:

[0038] The data acquisition module is used to collect microgrid operation data and upstream traffic flow data;

[0039] The model calculation module is used to build a traffic-energy potential field model and generate a net power demand forecast curve.

[0040] The thermal management control module is used to perform battery pre-temperature control based on the kinetic energy-thermal entropy mapping relationship;

[0041] The frequency decomposition module is used to adaptively adjust VMD parameters and decompose power commands based on the potential energy field gradient.

[0042] The power distribution module is used to collaboratively distribute power commands to supercapacitors and battery packs based on a virtual complex impedance model.

[0043] This invention provides a microgrid control method for highway service areas based on distributed photovoltaic energy storage. It has the following beneficial effects:

[0044] 1. Achieving a paradigm shift from "passive response" to "cross-domain prediction" in dispatching. This invention creatively constructs a traffic-energy potential field model, breaking through the limitations of traditional microgrids that rely solely on historical power data for trend extrapolation. By mapping the dynamics of traffic flow (density, speed, penetration rate) in the upstream physical space in real time to the impending energy surge pressure, the system can remotely sense potential charging demand tens of kilometers away. This cross-physical domain modeling method transforms the originally highly random and sudden electric vehicle fast-charging load into a quantifiable and traceable potential field strength, thereby securing a valuable pre-response time window for the microgrid dispatching system and fundamentally mitigating the voltage drop impact caused by pulsed high-power loads on weakly connected microgrids.

[0045] 2. Constructing a feedforward thermal safety defense line based on "cold energy buffer". Utilizing the implicit correlation between vehicle kinetic energy and battery charging rate, this invention proposes a kinetic energy-thermal entropy mapping mechanism. This mechanism keenly captures the characteristic of "high-speed driving often accompanied by low remaining battery power and high charging demand," identifying potential high heat generation risks before the vehicle even enters the charging station. By pre-regulating battery temperature during the window period before the load arrives, the system effectively constructs a "cold energy buffer" using the battery module's own thermal capacity. This feedforward strategy effectively avoids the thermal hysteresis effect of traditional temperature control systems during high-power fast charging, physically preventing battery temperature exceeding limits and thermal runaway.

[0046] 3. The frequency domain decomposition algorithm is endowed with "dynamic sensing" capabilities for precise current distribution. Addressing the complexity of power fluctuation frequency components, this invention abandons traditional filtering algorithms with fixed parameters. Instead, it utilizes the gradient of potential field intensity changes as prior information to adaptively adjust the variational mode decomposition parameters. When the system anticipates a severe power surge, it automatically adjusts the penalty factor to widen the decomposition bandwidth, ensuring that high-frequency surge components are precisely stripped and handled by the extremely fast-responding supercapacitor. This mechanism avoids the chemical battery being "passively" subjected to high-frequency ripple current due to frequency aliasing, ensuring that the "fast capacitor" and "slow battery" each perform their respective functions, significantly extending the overall lifespan of the hybrid energy storage system.

[0047] 4. Establishing a "Pain Avoidance" Power Self-Healing Mechanism Based on Virtual Impedance. By introducing a virtual complex impedance model that includes State of Health (SOH) and temperature deviation terms, this invention endows parallel battery packs with an adaptive self-protection capability. During microgrid operation, when a battery cell experiences a decline in health due to aging or an increase in temperature due to uneven heat dissipation, its virtual impedance will automatically increase. Based on the principle of parallel current sharing, the current will naturally avoid these "sub-healthy" cells and flow more towards the battery pack in better condition. This physical layer power allocation strategy, which does not require complex upper-level communication command intervention, effectively eliminates the circulating current problem between parallel batteries and avoids the "bottleneck effect" that accelerates overall system degradation.

[0048] 5. Robust closed-loop control with "self-evolution" characteristics. This scheme not only implements open-loop feedforward control but also integrates a closed-loop feedback mechanism of bus voltage monitoring and online parameter correction. The system can compare actual charging data of vehicles entering the station with predicted values ​​to optimize key parameters in the potential-power mapping function. This means that as the service area operates over time, the model can automatically adapt to seasonal changes, road condition changes, and load characteristic drift caused by the iteration of electric vehicle technology, ensuring that the control strategy maintains high-precision predictive capability and control stability throughout its entire lifecycle. Attached Figure Description

[0049] Figure 1 This is the main flowchart of the method of the present invention;

[0050] Figure 2 This is a hardware topology diagram of the microgrid control system for highway service areas according to the present invention;

[0051] Figure 3 This is a schematic diagram illustrating the modeling principle of the traffic-energy potential field model of the present invention;

[0052] Figure 4 This is a timing comparison diagram of the battery thermal management feedforward control based on kinetic energy-thermal entropy mapping according to the present invention;

[0053] Figure 5 This is a block diagram of the adaptive VMD decomposition logic based on the potential field gradient of the present invention.

[0054] Figure 6 This is a schematic diagram of the battery pack power collaborative distribution control principle based on virtual complex impedance according to the present invention. Detailed Implementation

[0055] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] Please see the appendix Figure 1 - Figure 6 This invention provides a method for regulating a highway service area microgrid based on distributed photovoltaic energy storage, including collecting operational status data, environmental data, and traffic flow data from the upstream road section of the service area microgrid. To support this implementation, this embodiment constructs a comprehensive physical architecture encompassing the energy supply side, energy storage regulation side, load demand side, and traffic sensing side. The service area microgrid adopts a DC bus coupled topology. Distributed photovoltaic power generation units are connected to the common DC bus through a maximum power point tracking controller, providing clean energy to the system. The hybrid energy storage system includes a supercapacitor bank and a lithium-ion battery bank, both connected in parallel to the DC bus through bidirectional DC / DC converters. The high power density of the supercapacitors is used to cope with instantaneous impacts, while the high energy density of the battery banks is used to handle continuous loads.

[0057] In the specific execution of data acquisition, the microgrid's operating status data... Data is obtained in real time through smart meters and battery management systems deployed on each branch line. At least the bus voltage at the point of common coupling Photovoltaic power generation Service area basic load power and the state of charge of supercapacitors in hybrid energy storage systems. and the state of charge of the battery pack Health status With internal real-time temperature Environmental data Data is obtained through micro-weather stations, primarily including light intensity. and ambient temperature It is used to assist in correcting the baseline parameters of photovoltaic power output prediction models and thermal management systems.

[0058] Traffic flow data for the upstream section of the service area In this embodiment, data is collected through roadside sensing units deployed along the main highway. These sensing units include, but are not limited to, millimeter-wave radar, high-definition cameras, or ETC gantry systems, and their coverage area is set to the distance from the service area entrance. to The upstream section of the road. The length of this section is set based on the travel time of electric vehicles at the current average speed to ensure that the forecast time window required for microgrid regulation can be covered.

[0059] The collected traffic flow data is processed by edge computing nodes to form a time-series vector containing multi-dimensional features. Specifically, the traffic flow data includes at least three core physical quantities: one of which is vehicle density. , indicating time In position The first is the number of vehicles per unit length, expressed in vehicles per kilometer; the second is the average vehicle speed. , indicating time In position The overall average speed of traffic flow, in kilometers per hour; the third is the electric vehicle penetration rate. This indicates the proportion of new energy vehicles in the current traffic flow that can be connected to the power grid for charging.

[0060] To eliminate the impact of sensor noise on subsequent modeling, this embodiment performs spatiotemporal synchronization and cleaning processing on the collected raw traffic flow data. This is particularly important for the average vehicle speed. The system uses a weighted average algorithm to smooth out abnormal speed-changing behavior of individual vehicles, thereby extracting a macroscopic velocity field reflecting the overall energy characteristics of the traffic flow. The acquired multi-source heterogeneous data is synchronously transmitted to the central control unit of the service area microgrid via a fiber optic ring network or a 5G private network, providing high-precision input variables for subsequent establishment of a traffic-energy potential field model and generation of net power demand prediction curves. All data is normalized before being fed into the model to eliminate calculation errors caused by different dimensions.

[0061] Based on the traffic flow data, a traffic-energy potential field model is established to calculate the potential field strength within a future time window. This model is then combined with microgrid operating status data and environmental data to generate a net power demand forecast curve for the microgrid. In this embodiment, the purpose of establishing this model is to transform the difficult-to-measure random charging behavior into a continuously differentiable physical field strength signal. Traditional power load forecasting is often based on autoregressive analysis of historical electricity consumption data, which is insufficient to address pulsed load scenarios like those in highway service areas, which are strongly coupled with upstream traffic conditions. This invention introduces the theory of physical potential fields to construct a virtual energy potential field. Electric vehicles traveling on upstream sections are equated to "particle streams" carrying energy demand. By integrating the motion states of these particle streams, the potential impact pressure on the service area microgrid is quantitatively assessed.

[0062] Specifically, this invention defines the time. Traffic-energy potential field strength This refers to the sum of the potential energy carried by all electric vehicles within a defined upstream range, which is about to be converted into charging load. This intensity depends not only on the number of vehicles but also, more fundamentally, on their state of motion. Physics dictates that vehicles traveling at high speeds consume electrical energy more quickly while maintaining a high kinetic energy state, and often have a lower state of charge (SOC) and a more urgent need for high-power charging when reaching a service area. Therefore, this embodiment defines the upstream road segment as a range... The time interval is calculated by weighting and integrating the kinetic energy corresponding to the vehicle density and average vehicle speed within the area. Traffic-energy potential field strength Its calculation expression is as follows:

[0063]

[0064] In the above formula, Indicates time Traffic-energy potential field strength, in joules equivalent; and These are the coordinates of the start and end points of the traffic flow monitoring section, which is located upstream of the service area. For a moment The electric vehicle penetration rate is used to eliminate the interference of fuel vehicles on the energy field; For a moment In position Vehicle density at the location; The normalized average vehicle mass coefficient, used to unify the dimensions; For a moment In position The average vehicle speed at a given location. This integral operation effectively compresses spatially distributed traffic flow information into a time-varying one-dimensional scalar signal. Its magnitude directly reflects the power surge pressure that will be exerted on the microgrid, and it is positively correlated with upstream vehicle density, electric vehicle penetration rate, and the square of the average vehicle speed.

[0065] To obtain a continuous potential energy field strength Subsequently, this embodiment further converts it into a dispatchable electrical power signal on the microgrid side. Due to the complex nonlinear conversion relationship between the potential energy field strength and the actual charging power (affected by factors such as charging pile power limitations and vehicle BMS protocols), this invention employs a data-driven method to construct a mapping mechanism. Using historical operating data from the service area, the potential energy-power mapping function is obtained through regression analysis or neural network fitting. The traffic-energy potential field strength Convert to future moment Electric vehicle charging load forecast ,in To predict the lead time, the average travel time of a vehicle from the monitoring point to the service area is used.

[0066] Finally, the system combines the regular electricity consumption and photovoltaic output of the service area to generate a net power demand forecast curve for the microgrid. This curve comprehensively reflects the energy exchange demand between the microgrid and the external power grid or energy storage system, and its calculation logic is shown in the following formula:

[0067]

[0068]

[0069] In the formula, For the future The predicted net power demand of the microgrid; when it is positive, it indicates that there is a power deficit that needs to be supplied by energy storage discharge or grid power. The base load forecast for the service area includes lighting, HVAC, etc. This invention predicts photovoltaic power generation based on environmental data. Through the above steps, the present invention successfully maps the physical fluctuations of upstream traffic flow to the power fluctuation prediction curve of the microgrid, providing a time-domain reference signal with an advanced perspective for subsequent energy storage thermal management feedforward control and refined power allocation.

[0070] Based on the kinetic energy characteristics of the vehicle's average driving speed, a kinetic energy-thermal entropy mapping relationship is constructed to predict the heat generation risk of the energy storage battery. Pre-temperature control of the energy storage battery's thermal management system is then implemented before the load corresponding to the net power demand prediction curve arrives. In this embodiment, this step aims to address the temperature exceedance problem caused by thermal hysteresis in high-power fast charging scenarios. Traditional battery thermal management systems typically operate based on real-time temperature feedback, meaning cooling is only initiated after the battery temperature actually rises. This passive response mechanism often fails to promptly curb the rapid temperature rise of the core due to thermal inertia when facing the pulsed, high-rate discharge conditions unique to highway service areas, leading to power derating or even thermal runaway risks.

[0071] This invention innovatively introduces the concept of "kinetic energy-thermal entropy mapping," utilizing the speed characteristics of upstream traffic flow as an early warning indicator of thermal risk. Physically, the speed of upstream vehicles does not directly translate into battery heat, but as a charging load intensity factor, it is closely related to the operating conditions the energy storage battery will face. Vehicles traveling at high speeds consume more energy per unit distance and typically have lower remaining charge (SOC) upon reaching a service area. This means they will maintain a longer constant-current fast-charging phase (CC phase) during charging, forcing the service area's energy storage battery to remain in a high-rate discharge state for an extended period. According to Joule's law, the heat generated inside the battery is proportional to the square of the current; continuous high-rate discharge will lead to a dramatic accumulation of heat. Therefore, this embodiment quantifies this mapping relationship from vehicle kinetic energy to accumulated battery heat by constructing a mathematical model.

[0072] Specifically, the average kinetic energy characteristics of the traffic flow are first obtained by calculating the square of the average speed of upstream vehicles. Subsequently, the kinetic energy-current-heat conversion coefficient was utilized. The average kinetic energy characteristics are mapped to the expected heat generation rate of the energy storage battery at future times. The expected heat production rate Based on the square of the vehicle's average driving speed and the battery's current internal resistance The calculation shows that it increases with the increase of the average vehicle speed, and the calculation model is expressed as follows:

[0073]

[0074] In the formula, This indicates the expected average heat generation rate of the energy storage battery in response to the upcoming traffic surge, expressed in watts. The kinetic energy-current-heat conversion coefficient is obtained through statistical regression of historical operating data and comprehensively reflects the vehicle's energy consumption characteristics and the efficiency of charging facilities. The average speed of upstream vehicles; This represents the current DC internal resistance of the energy storage battery, a value that is updated in real time based on the battery's state of equilibrium (SOH) and temperature. This formula reveals a positive correlation between high-speed traffic flow and the high risk of battery heat generation.

[0075] To achieve the expected heat production rate Subsequently, the control system further calculates the target reference temperature required to offset the thermal shock. The core idea of ​​this step is to utilize the battery module's own heat capacity as a "cold energy buffer," pre-lowering the battery temperature below the optimal operating point during the off-peak period before peak load arrives, thus reserving sufficient room for temperature rise. The target reference temperature... The calculation formula is as follows:

[0076]

[0077] In the formula, Set the target temperature for the pre-cooling operation of the thermal management system; This is the optimal electrochemical reaction temperature for the battery (usually around 25°C). This refers to the equivalent specific heat capacity of the battery module, expressed in joules per degree Celsius. For the predicted duration of high power load; This is the thermal safety margin factor, used to compensate for model errors and the effects of environmental heat leakage.

[0078] Based on the calculation Before the vehicle arrives at the service area Within the time window, the control system forcibly activates the thermal management system (such as a liquid chiller or air-cooled fan) to perform an advance action. The system adjusts the flow rate and temperature of the cooling medium to drive the actual temperature of the battery module. Towards Approaching. Through this feedforward pre-variable temperature control, when a high-power load is predicted... When actually applied to the energy storage battery, the heat generated inside the battery will first be used to fill the reserved temperature rise space, thereby ensuring that the battery's maximum temperature is always kept within the safe threshold throughout the entire discharge cycle, effectively avoiding lifespan degradation and safety hazards caused by high temperature.

[0079] The gradient of the potential field intensity is calculated, and the parameters of the variational mode decomposition algorithm are dynamically adjusted using this gradient to decompose the net power demand prediction curve into modal components of different frequencies. In this embodiment, this step aims to solve the power allocation problem of hybrid energy storage systems when dealing with non-stationary and nonlinear impact loads. Traditional filtering algorithms (such as low-pass filtering or wavelet decomposition) usually use fixed cutoff frequencies or basis functions, making it difficult to simultaneously ensure smoothness in steady state and response speed in transient state. Especially when facing the special load characteristics of highway service areas, which are "stable under normal conditions but sudden pulses," the decomposition with fixed parameters often results in the inability to effectively separate high-frequency impact components, causing the peak power that should be borne by the supercapacitor to "leak" to the battery side, thus damaging battery life.

[0080] This invention proposes a signal processing method based on prior knowledge of physical fields. The system utilizes the aforementioned traffic-energy potential field model to calculate the derivative of the potential field intensity with respect to time, thereby obtaining the gradient of the potential field change. The gradient value In a physical sense, this directly corresponds to the rate of change in power demand, i.e., the steepness of the load surge. When A larger value indicates an impending drastic power fluctuation; when A smaller gradient indicates a gradual change in load. The system utilizes this gradient information, which has predictive properties, to adaptively adjust the key parameter in the Variational Mode Decomposition (VMD) algorithm—the quadratic penalty factor. .

[0081] Specifically, secondary penalty factor This determines the bandwidth of the modal components (IMFs) obtained from VMD decomposition. The smaller the value, the wider the allowed modal bandwidth, and the richer the high-frequency mutation information can be captured; A larger value results in a narrower modal bandwidth and a smoother decomposition result, but it is also more prone to losing details. This embodiment constructs... and The adaptive adjustment function between them is expressed as follows:

[0082]

[0083]

[0084] In the formula, For a moment The gradient of the potential energy field change; For a moment The secondary penalty factor used; The baseline penalty factor corresponds to the optimal smoothing parameter under steady-state conditions. The gain coefficient is used to adjust the sensitivity of the gradient to parameter adjustments. According to this formula, when a gradient change in the potential energy field is detected... When the value increases, the control system automatically reduces the secondary penalty factor. The value of this adjustment widens the decomposition bandwidth, enabling the VMD algorithm to more accurately capture and separate the high-frequency impact components hidden in the prediction curve. Conversely, when the gradient of the potential field change decreases, the system increases. The numerical value enhances the algorithm's noise resistance and avoids misidentifying measurement noise as power commands.

[0085] After completing the adaptive parameter adjustment, the net power demand prediction curve is generated using the VMD algorithm. Decomposed into Individual intrinsic mode function (IMF) components Subsequently, the system sets the cutoff frequency based on the physical characteristics of the hybrid energy storage system. The frequency characteristics of each modal component obtained from the decomposition are evaluated, and they are recombined and divided into high-frequency power components. and low- and mid-frequency power section The division logic is as follows:

[0086]

[0087]

[0088] In the formula, For the first The center frequencies of each modal component. After the above processing, the previously mixed net power demand is precisely decoupled. This results in a device with high-frequency, large-amplitude fluctuation characteristics. Supercapacitor units, used as reference power commands, are allocated to the hybrid energy storage system to fully leverage their advantages of fast response and long cycle life, responsible for mitigating instantaneous surges; while those with relatively gradual changes and a higher proportion of energy components... This power is then allocated to the battery pack as a reference power command, leveraging its high energy density to handle energy transfer. This gradient-based dynamic frequency domain decomposition strategy ensures that the two energy storage media always operate in their respective most efficient frequency bands, achieving "dynamic decoupling" at the system level.

[0089] A virtual complex impedance model is constructed based on the health status and real-time temperature of the energy storage battery. Combined with the modal components, power demands at different frequencies are collaboratively allocated to the supercapacitors and battery packs in the hybrid energy storage system. After decoupling the total power into high-frequency and mid-to-low-frequency components in the aforementioned steps, this embodiment further addresses the power balance problem among multiple parallel branches within the battery pack. Due to variations in manufacturing processes and operating environments, inconsistencies in internal resistance, capacity, and state of aging (SOH) are inevitable among individual cells within a parallel battery pack. If current sharing control is simply implemented, "weak cells" with lower SOH will experience accelerated degradation due to overload, potentially even leading to thermal runaway. This results in a significant reduction in the effective capacity of the entire system due to the "weakest link effect."

[0090] This invention proposes an autonomous power allocation strategy based on virtual complex impedance. This strategy does not rely on complex upper-layer communication arbitration, but instead simulates the shunt characteristics in the physical circuit by introducing a virtual impedance term into the control loop of each parallel battery cell. The system targets each group of parallel battery cells in a hybrid energy storage system. Calculate their virtual impedance values ​​respectively. This virtual impedance value is not a fixed constant, but a dynamic function composed of the nominal impedance, the health status effect term, and the temperature deviation effect term. Its mathematical model is as follows:

[0091]

[0092] In the formula, For the first Virtual impedance setting value for battery cells; This is the nominal reference impedance of the battery pack, used to maintain basic droop control characteristics; This represents the current health status of the battery group (value range 0~1). The weighting coefficient is determined by the influence of health status. For real-time temperature monitoring of this battery group, For optimal operating temperature, The temperature deviation dead zone threshold (e.g., set to 5°C) means that temperature fluctuations within a certain small range are allowed to not affect the impedance. This is the temperature deviation penalty coefficient. The function ensures that impedance is increased only when the temperature deviation exceeds the dead zone.

[0093] Based on the above model, the virtual impedance value It exhibits specific dynamic characteristics: it varies with the battery health state. The value increases as the battery temperature decreases, and also increases with the real-time battery temperature. The resistance increases with the degree of deviation from the optimal temperature range (whether it is too cold or too hot). This design gives the battery cell a kind of "self-aware" physical property—the worse the condition, the greater the "resistance" it exhibits.

[0094] During the specific power allocation execution phase, the system utilizes the parallel circuit current splitting principle to allocate the low- and mid-frequency power components obtained from the aforementioned decomposition. Redistribute power. The power commands that each group of battery cells should handle. The virtual admittance is calculated based on the ratio of the reciprocal of its virtual impedance value (i.e., virtual admittance) to the sum of the total admittances, as shown in the following formula:

[0095]

[0096] In the formula, This represents the total number of parallel battery cell groups. Based on this allocation logic, those in healthy condition... Lower or temperature deviation Larger battery cells, because of their Large, the power received Naturally, smaller cells will have less power; conversely, cells in good condition will handle more power tasks.

[0097] This allocation mechanism achieves dynamic power balancing of the battery pack without manual intervention: it allows older, weaker batteries to recover and slows their aging process, while enabling younger, more powerful batteries to perform at their best. Ultimately, as operating time increases, the degradation rates of each battery cell tend to converge, thereby maximizing the overall lifespan of the hybrid energy storage system. Furthermore, by introducing a temperature penalty term, this strategy effectively suppresses the generation of localized hotspots, preventing systemic failures caused by overheating of individual batteries.

[0098] The method described in this embodiment also includes a closed-loop feedback correction step: real-time monitoring of the bus voltage deviation at the common connection point of the microgrid in the service area; when the bus voltage deviation exceeds a preset threshold, controlling the microgrid to exchange power with the main grid interconnection line to compensate for the deviation; and using actual vehicle data arriving at the service area to correct the parameters in the potential energy-power mapping function online. Although the aforementioned steps achieve high-precision feedforward prediction and control based on the traffic flow potential energy field, considering the uncontrollable factors in actual engineering such as random fluctuations in photovoltaic output and individual differences in vehicle charging behavior, the system must have the ability to compensate for prediction residuals in real time to ensure the voltage rigidity and operational stability of the microgrid.

[0099] In the specific implementation of closed-loop control, the system monitors the real-time voltage of the DC bus through a high-frequency sampling device. And calculate its relationship with the system's nominal rated voltage. instantaneous deviation between This deviation value is the most direct indicator of the power supply and demand balance in a microgrid. A voltage drop at the bus indicates a power deficit within the system; a voltage rise indicates a power surplus. The system has a set allowable voltage fluctuation dead zone threshold. ,when At that time, the system maintains the current energy storage scheduling strategy unchanged, relying on the passive voltage regulation characteristics of supercapacitors to absorb minor fluctuations.

[0100] When the bus voltage deviation is detected to exceed the preset threshold, i.e. This indicates that the power fluctuation amplitude has exceeded the preset regulation capacity of the hybrid energy storage system or that the prediction model has deviated significantly. At this point, the control system immediately activates the power exchange control logic with the main grid interconnection line to calculate the required compensation power. To achieve zero steady-state error tracking, this embodiment uses a proportional-integral (PI) control algorithm to generate reference commands, the calculation formula of which is as follows:

[0101]

[0102]

[0103] In the formula, This is the proportional gain coefficient, used for fast response to voltage surges; This is the integral gain coefficient, used to eliminate steady-state voltage errors. This is achieved by controlling the bidirectional AC / DC converter. The system can quickly absorb or feed power into the main grid, forcibly pulling the bus voltage back to a safe range, thus forming the system's last line of defense.

[0104] Meanwhile, to improve the prediction accuracy of the potential energy field model during long-term operation, this invention introduces an online parameter correction mechanism. During operation, the system records paired data between the "upstream traffic flow potential energy field intensity" and the "actual electric vehicle charging load." Once the vehicle enters the service area and completes the charging connection, the system acquires the actual charging power curve. And compare it with the power value predicted based on historical potential energy field strength. Compare the results and calculate the prediction error.

[0105] Using this prediction error, the system iteratively updates the key conversion coefficients in the potential-power mapping function using either gradient descent or least squares methods. It is assumed that the mapping function simplifies to a linear relationship. Then the coefficient The update rules can be expressed as follows:

[0106]

[0107]

[0108] In the formula, Represents the discrete correction time step. This refers to power prediction error; This refers to the time delay from upstream to the service area; The online learning rate determines the step size for parameter updates; These are the updated mapping coefficients.

[0109] Through the aforementioned correction process, the system can automatically adapt to environmental drift caused by seasonal changes (affecting air conditioning load and battery performance), changes in holiday travel patterns (affecting traffic flow characteristics), and the iteration of electric vehicle technology (affecting charging power demand). This self-evolving characteristic ensures that the traffic-energy potential field model always remains in an optimal fit, making the microgrid's control strategy increasingly precise with the accumulation of operating time, achieving a leap from "rule-based passive control" to "data-driven active learning."

[0110] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for microgrid control in highway service areas based on distributed photovoltaic energy storage, characterized in that, Includes the following steps: The system collects operational status data, environmental data, and traffic flow data of the service area microgrid and upstream road sections. The traffic flow data includes at least vehicle density, average vehicle speed, and electric vehicle penetration rate. Based on the traffic flow data, a traffic-energy potential field model is established to calculate the potential field intensity within the future time window, and a net power demand prediction curve for the microgrid is generated by combining the microgrid's operating status data and environmental data. Based on the kinetic energy characteristics of the vehicle's average driving speed, a kinetic energy-thermal entropy mapping relationship is constructed to predict the heat generation risk of the energy storage battery, and pre-temperature control is performed on the thermal management system of the energy storage battery before the load corresponding to the net power demand prediction curve arrives. The gradient of the potential field intensity is calculated, and the parameters of the variational mode decomposition algorithm are dynamically adjusted using the gradient to decompose the net power demand prediction curve into modal components of different frequencies. A virtual complex impedance model is constructed based on the health status and real-time temperature of the energy storage battery. Combined with the modal components, the power demand at different frequencies is collaboratively allocated to the supercapacitors and battery packs in the hybrid energy storage system.

2. The method for microgrid control in highway service areas based on distributed photovoltaic energy storage according to claim 1, characterized in that, The establishment of the traffic-energy potential field model and the calculation of the potential field intensity within a future time window include: The electric vehicles traveling upstream are represented in the model as an equivalent flow of particles carrying energy requirements; The traffic-energy potential field intensity at time t is calculated by weighted integral of the kinetic energy corresponding to the vehicle density and average vehicle speed within a set range of the upstream road section. The traffic-energy potential field strength characterizes the power surge pressure that will be exerted on the microgrid, and its value is positively correlated with the upstream vehicle density, electric vehicle penetration rate, and the square of the average vehicle speed.

3. The method for microgrid control in highway service areas based on distributed photovoltaic energy storage according to claim 2, characterized in that, The net power demand forecast curve for the generated microgrid includes: The potential energy-power mapping function obtained by fitting historical data is used to convert the traffic-energy potential energy field intensity into the predicted value of electric vehicle charging load. The net power demand forecast curve is obtained by adding the base load of the service area to the predicted electric vehicle charging load and then subtracting the photovoltaic power generation.

4. The method for regulating a highway service area microgrid based on distributed photovoltaic energy storage according to claims 1 to 3, characterized in that, The construction of the kinetic energy-thermal entropy mapping relationship based on the kinetic energy characteristics of the vehicle's average driving speed includes: The average kinetic energy characteristics of the traffic flow are obtained by calculating the square of the average speed of upstream vehicles. The average kinetic energy characteristics are mapped to the expected heat generation rate of the energy storage battery using the kinetic energy-current-heat conversion coefficient. The expected heat generation rate is positively correlated with the square of the vehicle's average driving speed and the battery's current internal resistance.

5. The method for regulating a highway service area microgrid based on distributed photovoltaic energy storage according to claim 4, characterized in that, The pre-temperature control of the thermal management system for the energy storage battery includes: Based on the expected heat generation rate and the specific heat capacity of the battery, calculate the target reference temperature required to offset the thermal shock; The target reference temperature is lower than the optimal electrochemical reaction temperature of the battery and decreases as the expected heat generation rate increases. Before the vehicle arrives at the service area, the thermal management system is activated in advance to adjust the battery temperature to the target reference temperature.

6. The method for microgrid control in highway service areas based on distributed photovoltaic energy storage according to claim 1, characterized in that, The method of dynamically adjusting the parameters of the variational mode decomposition algorithm using this changing gradient includes: The gradient of the potential field change is obtained by calculating the derivative of the potential field intensity with respect to time. The quadratic penalty factor of the variational mode decomposition is adaptively adjusted according to the gradient of the potential field change. When the gradient of the potential energy field changes increases, the value of the second-order penalty factor is decreased to broaden the decomposition frequency band and capture high-frequency impact components; when the gradient of the potential energy field changes decreases, the value of the second-order penalty factor is increased.

7. The method for microgrid control in highway service areas based on distributed photovoltaic energy storage according to claim 6, characterized in that, After decomposing the net power demand prediction curve into modal components of different frequencies, the method further includes: A cutoff frequency is set, and the modal components are divided into high-frequency power components and mid-to-low-frequency power components; The high-frequency power portion is used as a reference power and allocated to the supercapacitor in the hybrid energy storage system; The low- and medium-frequency power portion is used as a reference power and allocated to the battery pack in the hybrid energy storage system.

8. The method for regulating a highway service area microgrid based on distributed photovoltaic energy storage according to claim 7, characterized in that, The virtual complex impedance model constructed based on the health status and real-time temperature of the energy storage battery includes: For each group of parallel battery cells in the hybrid energy storage system, its virtual impedance value is calculated to allocate the low- and medium-frequency power portion. The virtual impedance value consists of the nominal impedance, the health status influence item, and the temperature deviation influence item; The virtual impedance value increases as the battery health status decreases, and also increases as the real-time battery temperature deviates from the optimal temperature range.

9. The method for regulating a highway service area microgrid based on distributed photovoltaic energy storage according to claim 8, characterized in that, The method of coordinating the allocation of power demands at different frequencies to the supercapacitors and battery packs in the hybrid energy storage system specifically involves allocating the low- and medium-frequency power portion to the battery packs, including: Using the principle of current division in parallel circuits, the power that each group of battery cells should bear is calculated based on the proportion of the reciprocal of the virtual impedance value of each group of battery cells to the sum of the reciprocals of the total reciprocals. This allows battery cells with lower health conditions or temperatures significantly deviating from their optimal range to exhibit high impedance characteristics, thereby enabling them to handle less power and achieving dynamic power balance among battery packs.

10. The method for regulating a highway service area microgrid based on distributed photovoltaic energy storage according to claim 3, characterized in that, The method also includes a closed-loop feedback correction step: Real-time monitoring of bus voltage deviation at the common connection point of the microgrid in the service area; When the bus voltage deviation exceeds a preset threshold, the microgrid is controlled to exchange power with the main grid interconnection line to compensate for the deviation. The parameters in the potential energy-power mapping function are corrected online using actual vehicle arrival data at the service area.

Images