An electric energy comprehensive management control method for high and low load working conditions

Abstract

The application relates to a kind of power comprehensive management control methods for high and low load conditions, belong to the field of energy supply system control.The method includes obtaining the current operation information of energy supply system, including working condition mode, target power demand and battery state of charge SOC value, and then judging the working condition category and calculating the system total power demand value.On this basis, according to the SOC value and power demand, the power distribution scheme of power generation unit and battery is determined: under normal working condition, the balance of power generation and battery charging and discharging is realized by setting the SOC threshold interval, and under emergency working condition, the maximum rate discharge of battery or the charging current is dynamically controlled to cope with load fluctuation.The system maintains the bus voltage and SOC in a safe range through closed-loop regulation, thereby improving the stability and response capability of the energy supply system.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electric energy management control, and particularly relates to an electric energy comprehensive management control method for high and low load working conditions. BACKGROUND

[0002] In the prior art, the electric energy management method for multi-source power supply systems has been widely applied to scenarios including electric vehicles, renewable energy systems and industrial equipment. Such methods usually obtain power demand, battery state and load change information, use fixed rules or model prediction strategies to allocate energy between power generation units and energy storage batteries to achieve stable and efficient system power supply. Some systems also introduce bus voltage monitoring and battery state of charge (SOC) limiting mechanisms to ensure operational safety.

[0003] However, existing electric energy management methods are mostly designed based on static working conditions or linear power response, and are difficult to adapt to frequent switching between high and low load operating states, especially lacking fine-tuning mechanisms in emergency load mutation scenarios. When load demand changes sharply or the power supply path is reconfigured, the system often has response lag, voltage overrun, battery overcurrent and other problems, making it difficult to balance system stability, battery life and load continuity, limiting its applicability in complex dynamic working conditions.

[0004] In view of this, it is necessary to provide an electric energy comprehensive management control method for high and low load working conditions to meet the needs of power supply coordination and safety control in a variable operating environment. SUMMARY

[0005] The present application provides an electric energy comprehensive management control method for high and low load working conditions to improve the stability and response capability of the power supply system.

[0006] The present application provides an electric energy comprehensive management control method for high and low load working conditions, characterized in that it comprises:

[0007] Obtaining the current operating information of the power supply system, including working condition mode information, target power demand and current battery state of charge SOC value;

[0008] Based on the current operating information, determine the current working condition category;

[0009] According to the working condition category and target power demand, calculate the corresponding system total power demand value;

[0010] determining the output power of the power generation unit according to a preset SOC threshold interval to meet the basic load demand, and the remaining power is balanced by charging and discharging of the battery within the SOC interval; when belonging to an emergency working condition, the battery is controlled to discharge at a maximum discharge rate when the load suddenly increases, the output power of the power generation unit is controlled to rapidly decrease when the load suddenly decreases, and the charging current of the battery is adjusted to avoid exceeding the maximum charging current that the battery can bear;

[0011] controlling the actual energy supply behavior of the power generation unit and the battery according to the output power of the power generation unit and the charging and discharging power of the battery;

[0012] detecting the system bus voltage and the battery state of charge SOC value based on the actual energy supply behavior, and if the bus voltage or the battery SOC value deviates from a set range, adjusting the output power of the power generation unit and the charging and discharging power of the battery according to the deviation closed loop until the bus voltage and the battery SOC value return to the allowable range.

[0013] The beneficial effects of the technical solutions provided in the application include:

[0014] (1) By dynamically identifying the working condition category and combining the target power demand and the battery SOC state, the organic coordination between the power generation unit and the battery is realized, which is suitable for complex operating environments with frequent switching of high and low loads, and significantly enhances the adaptability of the system to different operating conditions. (2) By real-time detection of the bus voltage and the SOC state, and establishment of a closed loop control mechanism based on deviation feedback, voltage overshoot or under-voltage caused by load fluctuation is effectively inhibited, and the stability and reliability of the energy supply process are ensured. (3) In an emergency working condition, by dynamically limiting the discharge rate and charging current of the battery, large current impact and overcharge and discharge problems are avoided, thereby reducing the risk of battery loss and improving the overall safety and battery life of the system. (4) By using a phased power regulation strategy, combining the SOC threshold and the power matching algorithm, the power generation unit main power supply capability and the battery regulation capability are complementary to each other, and the rapid response and regulation accuracy of the system under sudden power demand changes are improved. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 is an electric energy comprehensive management control method for high and low load working conditions provided by the first embodiment of the application. DETAILED DESCRIPTION

[0016] In the following description, many specific details are set forth in order to provide a thorough understanding of the application. However, the application can be practiced in many different ways beyond the specific details disclosed herein, and the skilled in the art can make similar modifications without departing from the spirit of the application, so the application is not limited to the specific implementations disclosed below.

[0017] The first embodiment of the present application provides a comprehensive power management and control method for high and low load conditions. Figure 1 , which is a schematic diagram of the first embodiment of this application. Figure 1 A first embodiment of the present application provides a comprehensive electric energy management and control method for use in high and low load conditions.

[0018] Step S101: obtaining current operating information of the energy supply system, including operating mode information, target power demand, and current battery state of charge (SOC) value.

[0019] In the process of implementing the comprehensive electric energy management and control method for high and low load conditions of the present invention, step S101, as the starting link of the entire control strategy, plays a key role in collecting system operating status data and establishing the basis for operating condition identification and power allocation.

[0020] When executing step S101, various operating data contained in the energy supply system should first be obtained. "Operating mode information" should include, but is not limited to, the current system operating stage, such as static standby, light-load operation, high-load operation, charging phase, or power surge phase. This information can be provided by the system scheduling module or upper-level controller via status words, mode codes, or preset operating condition tags, using a standardized data format for easy parsing and access.

[0021] Next, the "target power demand" data should be obtained. This data represents the system's power usage demand at the current moment or within a short-term forecast window, and is typically generated in real time by the load modeling module, scheduling module, or upper-level controller. The target power demand can be expressed as a real number (e.g., kW) per unit time, or as a discrete state level to identify the current load level, and transmitted to the power management control module via the data bus. To ensure timely response, it is recommended that the power demand signal have a fixed sampling period, preferably updated every 50 to 200 milliseconds to accommodate high-frequency load fluctuations.

[0022] Subsequently, the "current battery state of charge SOC value" should be obtained. SOC (State of Charge) is a state parameter for measuring the proportion of remaining battery capacity to full capacity, usually represented as a real value of 0% to 100%. When obtaining the SOC value, real-time measurement data provided by the battery management system (BMS) should be combined. The BMS usually calculates the SOC state value through sampling data of voltage, current, temperature signals, combined with extended Kalman filtering, Coulomb counting method, etc. When the system calls the SOC value, it needs to ensure that the value is the updated data in the latest period and has been filtered to reduce misjudgment caused by instantaneous noise. To further improve the reliability of the SOC value, a SOC change rate constraint judgment mechanism can be introduced in the design to prevent abnormal jumps.

[0023] After the acquisition of the three information parameters is completed, a structured data packet should be constructed in the management system for subsequent step calling. The data packet should at least include the working condition mode code, the target power value, the current SOC value, and the corresponding time stamp or data validity identifier. It is recommended that the packaged data be uniformly stored in the controller cache area and a shared memory channel be established between the judgment module to ensure the real-time and integrity of data access.

[0024] Therefore, the implementation of step S101 should include multiple sub-operation processes such as multi-source data acquisition, numerical analysis, logical judgment, data packaging, and shared cache establishment. Each step should ensure accurate sampling, timely refreshing, and clear structure to lay a foundation for subsequent working condition discrimination, power calculation, and control execution. Only by ensuring the accuracy and integrity of this step implementation can the response speed and control precision of the entire electric energy comprehensive management system reach the ideal state.

[0025] Further, the current running information of the energy supply system includes:

[0026] Based on the task instructions or running control instructions received by the energy supply system, the working condition label corresponding to the working condition mode information is identified by analyzing the working stage, driving state, and environmental adaptation strategy implied therein, and the current working condition mode information is extracted and generated;

[0027] After obtaining the working condition mode information, the reference power data corresponding to the working condition label in the historical load model and environmental database are called, and the target power demand required at the current time is deduced and output by combining the real-time measured load current and voltage change trend;

[0028] After the target power demand is determined, the battery voltage, current, and temperature data fed back by the battery management system are read in real time, the remaining battery capacity is estimated by the Coulomb integral method after temperature compensation correction, and the current battery state of charge SOC value is calculated and output;

[0029] Based on the current target power demand and the SOC value, it is judged whether the battery can complete the power supply demand without exceeding the safe operation boundary, and the judgment result is compared with the aforementioned working condition mode information to determine whether there is a power supply-demand mismatch risk.

[0030] Firstly, the process of obtaining the current running information of the power supply system starts with the reception of task instructions or running control instructions. These instructions are usually issued by the upper dispatching system, task management platform or human-computer interaction interface, and contain the task type, expected duration, task priority, target route or task path, required load level and coordination requirements with the ground or environment of the current power supply system. The instruction is transmitted in a predefined data format, and the system reads and extracts the required fields through the built-in protocol parser, and further matches these fields with the preset working condition definition rules. For example, when the task instruction contains keywords such as "point hovering", "vertical take-off and landing" or "long-distance cruising", the system determines that it is currently in the corresponding working condition label, such as silent flight, take-off climb or stable cruising state. In some scenarios, environmental adaptation strategies such as "high wind speed adjustment", "extreme temperature warning" or "complex terrain obstacle avoidance" are embedded in the task instructions to activate special running strategies to improve the robustness and response sensitivity of the system.

[0031] After identifying the working condition mode information, the system will enter the target power demand generation stage. This stage relies on a joint reference knowledge base composed of historical load models and environmental databases. The historical load model is constructed through the statistical and modeling of long-term running data. After a large number of field tasks are accumulated during the system trial running stage, the current, voltage, load response time and power consumption parameters under multiple typical working conditions are normalized and sample clustering analysis is performed to form a mathematical description set. This model can be constructed using methods such as multi-dimensional linear regression, K-means clustering or BP neural network, and is divided into sub-models according to different environmental factors such as temperature, humidity, air pressure, altitude, wind speed, etc. to improve the power prediction accuracy. The environmental database records the environmental perception information at different space-time coordinates, and the data sources include airborne environmental sensors, ground remote sensing systems and historical records. After identifying the current working condition label, the system automatically retrieves the corresponding historical load model and its power offset correction factor under the current environmental conditions, and combines the current measured real-time load current and voltage change trend to derive the target power demand value at the current time using differential prediction or short-time window sliding weighted average algorithms. This value is not only used to drive the subsequent energy allocation strategy, but also can realize front-end power preparation in the multi-module collaborative power supply scene.

[0032] After the target power demand value is determined, in order to ensure the safety and efficiency of energy supply, the current battery state of charge SOC value needs to be accurately obtained. This process is based on the battery management system (BMS) to read the voltage, current, temperature and charging / discharging state information of the battery in real time. Considering that the SOC value of lithium battery is significantly affected by temperature, the system needs to introduce a temperature compensation mechanism, that is, based on the original data of BMS, the voltage is corrected by using a temperature sensitive coefficient according to the current environmental temperature or the surface temperature of the battery cell, and then the SOC is calculated. The SOC estimation method can use the improved coulomb integral method, that is, the total power change is recorded from the initial state of battery charging and discharging, and the drift error of current sensor, sampling interval offset and integral window updating strategy are considered in the calculation process, and finally the current battery SOC estimation value is output by combining the open circuit voltage and the corrected result of the experience curve. In order to further improve the estimation accuracy, the system can use a fusion method, such as Kalman filtering fusion of model prediction value and sensor measurement value, so as to realize dynamic correction.

[0033] Finally, in order to form the early warning and evaluation ability of the system supply and demand risk, the system compares and analyzes the target power demand and the current SOC value. Here, the key is to judge whether the battery has the ability to continuously supply power within the safe range to the next task node under the current state of charge. The system compares the current SOC with the minimum SOC threshold required by the task, and investigates the load power gradient change in the future scheduling period, to predict whether the battery discharge depth is likely to exceed the recommended range. If there is obvious risk, further classification judgment is made combined with the working condition mode information, for example, if the current working condition is emergency obstacle avoidance and the SOC is about to be lower than 10%, the system will report the risk in advance and suggest entering the energy saving mode or forced load reduction mode.

[0034] Step S102: determining the current working condition category based on the current running information.

[0035] When implementing the power comprehensive management control method for high and low load working conditions according to the present application, the role of step S102 is to determine the working condition category of the power supply system based on the current running information after obtaining the information, and to provide a clear basis for subsequent power demand calculation and energy distribution strategy selection.

[0036] Specifically, the classification of working condition categories should be based on the comprehensive judgment of the current working mode, load state and dynamic characteristics of power demand of the system. Generally, the working conditions can be divided into two categories: normal working condition and emergency working condition. The normal working condition includes the running of the power supply system under the condition of stable power demand without sudden state, such as regular driving, stable power supply, regular charging, etc.; while the emergency working condition covers abnormal states such as sudden change of system load, battery state close to limit, failure of power redundancy, etc. This kind of working condition is usually accompanied by characteristics such as rapid rise or fall of power, severe voltage fluctuation, etc.

[0037] To automatically determine the operating condition, the controller performs a joint analysis of the three types of operating information obtained in step S101. First, based on the operating tag or status code indicated in the operating mode information, it determines whether the system is currently in a mode that could trigger a load change, such as transitioning from standby to operation or from normal operation to power compensation. This determination can be implemented using a pre-set state transition table, a finite state machine (FSM), or threshold judgment logic.

[0038] Secondly, based on the changing trend of the target power demand, we determine whether the current power request exhibits a sudden change. For example, if the power request increment ΔP per unit time exceeds a preset power surge threshold, or if the power request decreases below a preset power decrease threshold, we can preliminarily determine that the system has entered an emergency load state. This type of judgment can be achieved by performing real-time differential calculations on the power request curve using a sliding window approach, combined with delay filtering to avoid misjudgments.

[0039] Finally, the battery's state of charge (SOC) value plays a crucial role in determining operating conditions. When the SOC value approaches the warning threshold, for example, below 30% or above 90%, and coincides with a sudden power surge, it indicates that the system's energy storage buffering capacity is insufficient. In this case, the current state can be classified as an emergency operating condition, triggering a more conservative charge and discharge control strategy. A hysteresis mechanism can also be implemented to prevent control oscillations caused by frequent switching between operating conditions.

[0040] Ultimately, by logically synthesizing the above three types of judgment results, a stable and accurate operating condition category output can be generated. Preferably, the controller should establish a complete operating condition mapping table or decision tree, mapping different information combinations to clear operating condition identifiers, such as "normal-steady state," "normal-charging," "emergency-sudden increase," and "emergency-sudden decrease." This will drive the strategy branch selection of the subsequent power demand calculation module.

[0041] Therefore, step S102 is not just a static judgment operation, but a dynamic classification process that integrates pattern recognition, trend prediction, and state reasoning. Its accuracy directly determines the response rationality and operational stability of the entire electric energy integrated management and control system. In engineering implementation, it is best to avoid relying solely on a single variable for classification judgment. Instead, the operating condition identification module should be constructed based on a comprehensive analysis of multi-dimensional operating information to ensure that the system has good operating condition adaptability and abnormality identification robustness.

[0042] Furthermore, the operating condition categories include normal operating conditions and emergency operating conditions, wherein the normal operating conditions include parking charging, silent driving, land charging, vertical take-off and landing, and cruise flight, and the emergency operating conditions include emergency braking, high-power maneuvering, fan damage, and emergency obstacle crossing.

[0043] In the implementation of the proposed power comprehensive management control method for high and low load conditions, the classification of working conditions needs to be accurately identified and responded to according to the actual operating state of the system. Therefore, in combination with the operating mode of the aircraft or ground vehicle, the working condition categories are clearly distinguished into two basic dimensions of normal working condition and emergency working condition, and each category is further subdivided into several typical scenarios to meet the response needs of power management under different load conditions.

[0044] In normal working conditions, the system is usually in a steady state or regular operating state, and there is no sudden disturbance or severe load fluctuation. Specifically, parked charging refers to the process of charging the battery through external charging equipment when the system is in a stationary state, at which time the load is small and constant, and the system mainly focuses on the accurate control of charging power and the safe upper limit management of SOC. Silent driving refers to the slow movement of a vehicle or aircraft at very low power, which is usually used in silent working conditions during personnel or material transportation, and has lower requirements for transient response capability of energy, but has higher requirements for continuity of smooth energy supply. Road charging is a running and charging working condition, which is commonly seen in hybrid or extended range systems. In such working conditions, the system needs to coordinate the power distribution between engine output and battery charging, so that the power supply and vehicle operation do not interfere with each other. Vertical take-off and landing is a typical high-load starting process specific to aircraft, at which time the propeller or propulsion unit needs to provide a large amount of thrust in a short time, which puts high requirements on the instantaneous high-power output capability of the power supply system. Cruise flight is the stable stage of flight, and the load is relatively stable, and the system needs to maintain constant power output and timely charge or recover energy from the battery to optimize energy efficiency.

[0045] In emergency working conditions, the system must respond to sudden events or temporary high-risk operations, and needs to have the ability to quickly adjust the power supply strategy to ensure safe operation. Emergency braking often occurs during road travel, and the system needs to quickly reduce the output power and charge the battery according to the needs of the feedback type, at which time the timeliness and accuracy of power regulation response are extremely high. High-power maneuvering usually accompanies rapid acceleration, sharp turning or special flight attitude adjustment process, which requires the power supply system to output abnormal power in a short time, and the discharge rate scheduling strategy of the battery becomes the core of control at this time. Fan damage is a typical emergency working condition of hardware failure, and the system needs to quickly reduce the overall load after detecting abnormal speed, sudden current change or temperature rise, etc. to avoid fault propagation, and start the redundant power supply path or forced load shedding mechanism. Emergency obstacle avoidance is a high-risk working condition triggered in temporary avoidance operations in path-limited or complex terrain areas, which requires the system to mobilize all available power resources to achieve high-power burst power supply in a short time, and cooperate with a rapid power drop process to prevent system overheating or battery overdischarge caused by continuous high-power output.

[0046] In summary, the setting of operating condition categories is the basis for this method to achieve intelligent energy control. In practical applications, it is necessary to combine operating data, the current mode of the vehicle / aircraft, and the battery status for real-time judgment and classification to ensure the refined control and scheduling of the energy supply behavior between the power generation unit and the battery under different operating conditions, ultimately improving the stability, safety, and energy efficiency of the overall operation of the system.

[0047] Furthermore, determining the current operating condition category based on the current operating information includes:

[0048] Based on the operation target semantic fields, path planning labels and dynamic environment intervention items extracted from the operation mode information, a task context semantic vector is constructed and mapped to a multidimensional operation condition label embedding space;

[0049] In the multi-dimensional operating condition label embedding space, a dynamic clustering algorithm is used to identify the closest historical operating condition cluster center, and combined with the current target power demand change trend, a preliminary operating condition classification result and a corresponding credibility score are output;

[0050] Based on the preliminary operating condition classification results, the dynamic energy consumption curve family and the charge safety limit parameter group associated therewith are called to construct a safety mapping space of the battery state of charge (SOC) value, and the current SOC value position is combined with the determination of whether it is in the controllable load window;

[0051] If it is determined to be at the edge of the load window, the real-time environmental disturbance index is integrated and matched with the historical working condition anti-disturbance model, the original working condition classification result is corrected and the working condition category priority is reordered, wherein the real-time environmental disturbance index includes sudden change in wind speed, sudden drop in temperature, and electromagnetic interference;

[0052] The operating condition category corresponding to the highest confidence level in the comprehensive matching results is determined as the current operating condition category.

[0053] First, after obtaining the working condition mode information, the system needs to further extract the semantic fields contained therein. Specifically, it includes the work target field, path planning label and environmental intervention description item. For example, the work target field may correspond to instructions such as "completing a fixed-point route" and "continuous high-power vertical hovering", and the path label covers motion elements such as "straight line", "turning", and "climbing". Environmental intervention items refer to external events such as "crosswind influence" and "strong electric field interference". In order to characterize the above heterogeneous information, the system introduces a task semantics-oriented embedding method, which encodes the semantic field into a multi-dimensional embedding vector and combines it into a single task context semantic vector through an attention weighting strategy. This vector is projected into the working condition label embedding space as input. This space is a predefined multi-dimensional vector space. Each historical working condition category is encoded as a fixed label vector to form the reference cluster center.

[0054] Subsequently, a clustering matching algorithm based on dynamic cosine similarity is used for preliminary working condition classification. The algorithm calculates the distance between the current constructed task semantic vector and the historical working condition label cluster center according to the similarity, and at the same time, the change rate of the target power demand is used as a dynamic weighting factor to intervene in the clustering process, so as to improve the sensitivity of the classification result to real-time power fluctuations. The output is the preliminary determined working condition category and its corresponding confidence score. The scoring mechanism adjusts the decision boundary according to the standard deviation of the similarity distribution, ensuring that the confidence has a numerical interpretation.

[0055] Based on the preliminary working condition classification result, the system further queries the dynamic energy consumption curve family and state of charge limit parameters associated with the working condition. The dynamic energy consumption curve family is obtained by training the measured data under historical tasks, describing the nonlinear response characteristics of the power load to the SOC value decrease per unit time, and the state of charge limit parameters include SOC threshold interval, maximum charge and discharge current capacity limit and other information. After substituting the current SOC value into the above model, the system constructs a two-dimensional safety mapping space with SOC as the horizontal axis and load stability as the vertical axis, to determine whether the current SOC value is in the load controllable window, i.e. whether the system still has enough buffer to respond to load disturbance.

[0056] When it is determined that the SOC value is at the edge of the controllable window, i.e. about to exceed or just enter the safety interval, the system further retrieves real-time environmental disturbance indicators and fuses them with historical working condition disturbance resistance models. Real-time disturbance indicators can include wind speed variation amplitude, temperature sharp drop, electromagnetic interference frequency, etc., while historical disturbance resistance models are based on prior experience or measured data to establish a mapping relationship between environmental disturbance intensity and power compensation failure probability. The fusion of the two is performed through a residual convolutional network to interact features, fine-tune the working condition classification confidence, correct the original sorting result and rearrange the working condition priority, so that the final output has better dynamic adaptability.

[0057] Finally, by integrating all matching and correction results, the system selects the working condition category with the highest confidence as the current working condition type, and enters the subsequent power allocation and energy supply control stage. It should be noted that this method does not use the traditional static mode based on fixed logic judgment in the working condition discrimination process, but through the construction of semantic vector space and environmental disturbance fusion channel, it realizes high adaptability and scalability, especially suitable for real-time energy supply decision-making scenarios of new flight platforms, electric vehicles or complex terrain operation equipment under different load changes.

[0058] This embodiment proposes and introduces a working condition history disturbance resistance model. The core purpose of this model is to provide a quantifiable, comparable, and decision-making auxiliary judgment mechanism for different external disturbance situations faced by the system, such as sudden changes in wind speed, dramatic temperature drops, or electromagnetic interference, to further adjust the working condition classification results based on the changes in disturbance conditions, and improve the adaptability and robustness of the control strategy on the basis of the preliminary identification of the current working condition category.

[0059] The construction process of this model is as follows:

[0060] First, during the initial deployment or operation of the system, design to collect and record historical task data under multiple typical working conditions. Specifically, for working conditions such as "cruise flight", "vertical take-off and landing", "parking charging" and others, continuously collect external disturbance indicators during their operation period, such as the instantaneous change amplitude of wind speed, the rapid drop rate of ambient temperature, the signal-to-noise ratio attenuation amplitude of wireless communication under electromagnetic interference, etc. At the same time, the dynamic response behavior of the system after these disturbances need to be recorded, including but not limited to the offset of battery voltage, the fluctuation amplitude of power output, the time required for the system bus voltage to recover to a stable state, etc.

[0061] Next, for each type of working condition, its response behavior under various disturbance conditions is archived and organized to form a working condition disturbance response file. For example, in the case of a rapid increase in wind speed to 20 meters per second, the battery voltage under a certain working condition may drop by 0.5 volts within 5 seconds and recover to the normal fluctuation range after 8 seconds; while in another working condition, it may take 15 seconds or even longer to recover. This information will be organized into a set of data records to describe the "disturbance resistance performance" of the working condition under this type of disturbance.

[0062] Through statistical analysis of the same working condition under different disturbance conditions in a large number of historical tasks, the disturbance resistance characteristics of each type of working condition under various disturbance dimensions can be summarized. These characteristics can be represented by qualitative description items, such as: "high adaptability" to wind speed mutation, "moderate adaptability" to temperature variation, and "low adaptability" to electromagnetic interference. Or through a normalized scoring standard, convert recovery time, fluctuation amplitude, system offset degree, etc. into a unified adaptability level (such as a score between 0 and 1), to build a working condition disturbance resistance characteristic file.

[0063] When the system enters the running phase, if it detects that the current disturbance environment has changed significantly, such as sudden increase in wind speed, sudden drop in ambient temperature, or unstable wireless communication state, it will immediately extract the actual observation value of the current disturbance indicator and match it with the working condition disturbance resistance file. The matching method uses "similarity comparison", that is, to determine which disturbance conditions of the working conditions in history are closest to the current environment disturbance, and at the same time, to refer to the system performance of these working conditions under similar environments, to speculate whether the current preliminary working condition category has sufficient disturbance resistance ability.

[0064] For example, the system initially identifies the operating condition as "silent driving," but in real time detects a rapid increase in wind speed from 5 meters per second to 20 meters per second. Historical records show that "silent driving" frequently experiences excessive voltage fluctuations under these wind speed conditions, while "cruise flight" maintains stable battery voltage under similar disturbances. Therefore, the system will combine current environmental disturbance data, the changing trend of the target power demand, and whether the battery state of charge is approaching its limit to determine whether the current operating condition identification result needs to be revised, switching to a more resistant operating condition category and deploying a more appropriate energy allocation strategy in advance.

[0065] The entire disturbance rejection model is adaptively updated as mission data accumulates. For example, after each mission, the system reassesses the actual disturbance conditions and system performance, adding this new data to the operating profile for more accurate disturbance rejection matching analysis in future missions.

[0066] The following provides a detailed example:

[0067] During a vertical takeoff and landing (VTOL) unmanned flight mission, the energy supply system received the following mission instruction: "Vertical takeoff and landing mission, target area radius 15 meters, expected hovering time 2 minutes, ambient wind speed warning level 2, path set." In this example, the energy supply system first extracts the key elements of the mission instruction based on the original mission data issued by the mission dispatch center using a preset semantic parsing model, including:

[0068] Operation target semantic fields: vertical take-off and landing, hovering

[0069] Path planning tags: target area radius, fixed path

[0070] Dynamic environmental intervention item: Wind speed warning level 2

[0071] The above fields are constructed into a contextual semantic vector. The construction method is: each semantic keyword is encoded into a sparse feature vector, embedded through Word2Vec model pre-training, and then the final task semantic vector is obtained by weighted average. Assumptions:

[0072] ("Vertical Take-off and Landing", "Hanging", "Target Area 15 meters", "Wind Speed ​​Warning Level 2")

[0073] Output a dimension of The vector of .

[0074] Then, the system maintains a multi-dimensional working condition label embedding space , which predefines the embedding vectors and cluster centers of all historical working condition labels. For example:

[0075] : Parking charging center vector

[0076] : Silent driving center vector

[0077] : vertical take-off and landing center vector

[0078] Through the dynamic Kmeans clustering algorithm, the system can identify To which cluster center it is closest, the cosine similarity is used for calculation:

[0079]

[0080] in, Represents the current task context semantic vector. This vector is a multidimensional numerical vector obtained by feature extraction and embedding coding of input operation instructions, path planning, environmental intervention and other information, usually located in the real vector space Its dimensions It depends on the output design of the semantic feature encoder, for example, it can be set to 128 dimensions or 256 dimensions. Indicates the The semantic center vector corresponding to the historical working condition category label. The “representative” semantic vector representing a certain type of working condition (such as vertical take-off and landing, cruise flight, etc.) in history is formed by clustering or averaging in multiple missions and is used to construct the cluster center in the embedding space. Its dimension is the same as Stay consistent.

[0081] Assume that at this time (Vertical take-off and landing) has the highest similarity, and the output preliminary working condition classification result is vertical take-off and landing, and its credibility score is set to 0.94.

[0082] To further improve judgment accuracy, the system also performs cross-validation based on the current target power demand trend. The system monitors the flight control system's load current and power fluctuation rate in real time. Assuming a power growth rate of 2.1 kW per second over the past 10 seconds, it identifies a non-resident operation and supports the initial determination of vertical takeoff and landing.

[0083] Next, the system reads the SOC value provided by the battery management system , current temperature , the real-time voltage is 51.2 V and the current is 4.5 A. Through the temperature compensation coefficient (correspond The current is calculated as follows:

[0084]

[0085] where, is the current state of charge; represents the actual remaining capacity of the battery at time point as a percentage of its rated capacity. For example, = 62% indicates that the current remaining capacity is 62% of the battery's rated capacity. This value is a key basis for the battery control system to make power supply strategy decisions. is the SOC value of the battery at time point . It is usually obtained from the previous round of measurement or system record, and is the starting point for the current SOC calculation. is a coefficient determined according to the current battery temperature , reflecting the influence of temperature on the charging and discharging efficiency of the battery. When the temperature deviates from room temperature (e.g. 25°C), the discharging efficiency of the battery will decrease, thus the current current estimation value needs to be adjusted. For example, at 25°C, the empirical value is set as κ(25°C)=0.96, indicating that the actual effect of current at this temperature is 96% of the theoretical value. This coefficient can be obtained by fitting the efficiency-temperature curve provided by the battery manufacturer, or by constructing a lookup table through experimental data.

[0086] is the current value of the battery at time , with the unit of amperes (A). If the discharge is constant, this value is a constant (e.g. 4.5 A in the example).

[0087] is the rated capacity of the battery, with the unit of ampere-hours (Ah), which is the maximum total charge that the battery can provide at a normal discharge rate. For example, 10 Ah indicates that the battery can sustain a 1 A current for 10 hours.

[0088] The estimated available capacity of the battery is approximately 6.2 Ah, which meets the current total demand power of about 0.126 kWh for 10 seconds. Then, the system loads the dynamic energy consumption curve family associated with the "vertical take-off and landing" label and the SOC limit parameter set, for example, it is stipulated that the minimum SOC for "vertical take-off and landing" operation should be , and the maximum is . The current is in the safety window, but the system further detects the environmental disturbance item "sudden increase in wind speed", with a value of 3.7 , which exceeds the secondary warning threshold in the local model.

[0089] To this end, the system calls the environmental immunity fusion model to match the disturbance indicators with the historical immunity vectors of the working conditions, and calculates the current working condition stability value as 0.79 through the similarity re-scoring mechanism, which is slightly lower than the default safety threshold of 0.85, prompting the system to correct the working condition priority. In this context, "silent idle" is promoted to the sub-optimal working condition category.

[0090] Finally, the system still determines "vertical take-off and landing" as the current working condition category after evaluating the credibility, but with a disturbance monitoring mark, prompting that redundant power should be dynamically reserved in the subsequent control strategy.

[0091] Step S103: Calculate the corresponding total power demand value of the system according to the working condition category and the target power demand.

[0092] In the execution of step S103, the system calculates the total power demand value required by the energy supply system at the current time according to the determined working condition category and the data obtained from the target power demand. The core of this process is to comprehensively consider the running state of the current load, the response ability of the power supply system and the participation ability of the energy storage device, so as to form a reasonable and dynamically adjustable power demand benchmark for guiding the subsequent power distribution process.

[0093] Firstly, the system sets a basic framework for the entire calculation process according to the working condition category obtained in step S102. If the current system is in normal working condition, it means that the load demand is relatively stable and predictable, so the target power demand can be directly used as the main reference value. For example, if the target power demand reflects the average power consumption of the current load device during normal operation, the system can use this value as the main body of the total power demand. On this basis, the system also needs to consider a certain operating margin to cope with possible slight load fluctuations in a short time. This margin can be set as a fixed proportion based on experience, or automatically estimated by analyzing historical operation data.

[0094] If the current working condition is determined to be an emergency state, the calculation logic will be more conservative. The system needs to determine whether there is a rapid increase trend of the load, such as the sudden start of some critical load, load transfer or the instantaneous pressure brought by the failure of the upper-level power supply. At this time, the system needs to artificially increase a reasonable safety redundancy while referring to the target power demand, reserving the dynamic adjustment space of the energy supply system. This redundancy can be obtained by comprehensively evaluating the working condition level and the load response curve, and its size depends on the growth rate, duration of the sudden load and the response ability of the battery and the power generation unit.

[0095] In addition to the above considerations for the load itself, the current state of charge of the energy storage device must also be considered when calculating the total power demand value of the system. If the battery's state of charge is within a reasonable range, that is, not close to the low or high limit, the system assumes that the battery has normal discharge or charging capabilities at the current moment and can be added to the total energy supply framework as a power balancer. However, if the battery SOC is close to the edge of the allowable range, for example, below the minimum safe value, the system should reduce its reliance on the battery's energy supply capacity and use the power generation unit as the main energy supply object, thereby maintaining the continuity of the system's power supply while ensuring battery safety.

[0096] Furthermore, power demand calculations must consider the characteristics of each load in the system. For example, if there are high-impact loads or intermittent loads, the system should identify these loads based on past operating data and pre-set characteristic templates, and account for dynamic power demands during startup or changes in the power calculation. This identification can be achieved by the controller monitoring the power change slope or by using feedforward predictions based on the load access plan provided by the upper-level task scheduling system.

[0097] After the calculation is complete, the system generates a comprehensive output containing information such as actual power demand, safety margins, and energy storage availability assessments, which is then fed into the subsequent power allocation logic. This output should be in a structured data format that supports real-time sharing and rapid access across different submodules, ensuring consistency, stability, and traceability of the overall control process under all operating conditions.

[0098] Therefore, the implementation of step S103 is not only a process of generating power figures, but also a multi-dimensional comprehensive judgment of load status, power supply capacity, system safety and dynamic response. The key lies in providing a power demand benchmark with strong adaptability and high control accuracy through working condition drive, information fusion and dynamic adjustment mechanism, thereby supporting subsequent energy management decisions.

[0099] Furthermore, the calculating of the corresponding total system power demand value according to the operating condition category and the target power demand includes:

[0100] Combined with the current operating condition category, a power demand record sequence matching the operating condition is extracted from historical data. Based on the power value changes in each period in the record, the average power level, maximum power surge amplitude, and power fluctuation duration at different stages are statistically analyzed to form a basic load characteristic description.

[0101] Based on the task type, execution phase, and running time under the current working conditions, the target power demand is subdivided into multiple continuous time periods, and a corresponding load level is assigned to each time period. Then, the load level is matched with the corresponding power parameters in the basic load characteristic description to determine the preliminary target power demand sequence;

[0102] Real-time collection of influencing factors related to the current operating environment, including wind speed, temperature, humidity, terrain undulation, and obstacle density. Based on the impact of different environmental factors on power consumption, the power value of each period in the preliminary target power demand sequence is adjusted step by step to obtain a target power sequence that includes environmental correction factors.

[0103] Compare the current battery state of charge with the power buffer to assess whether the battery has the ability to handle a sudden load surge or release part of the base load. If the conditions are met, a corresponding load transfer correction value is introduced into the target power sequence to match the target power with the battery's response capability.

[0104] The adjusted target power values ​​for each time period are combined with the basic power demand to form a set of final system total power demand values ​​in chronological order, which are used for the subsequent execution of power allocation and energy control strategies, so that the calculation results can fully reflect the comprehensive effects of working condition history, mission characteristics, environmental impact and battery capacity.

[0105] In actual integrated power management scenarios, systems often face complex factors such as alternating high and low loads, sudden load shocks, and uncertain external environmental interference. Traditional approaches that rely on static power mapping or simple task power tables are difficult to adapt to dynamically changing operating conditions. To this end, this embodiment proposes a target power calculation strategy that integrates multiple factors and refines them in multiple stages. This strategy aims to achieve highly timely and adaptable total system power demand forecasts, and provide an accurate reference for subsequent coordinated power supply control strategies between power generation units and batteries.

[0106] First, the power demand record sequences archived in the historical database are filtered based on the currently identified operating condition category. Specifically, each historical record sample contains multiple record fields corresponding to the operating state, such as task type, phase description, timestamp, real-time load power, and battery participation ratio. Based on a matching mechanism for operating condition categories, the system extracts a collection of historical samples with similar tags and statistically processes the power values ​​from different time periods within this collection to generate a typical load profile for that operating condition category. This profile includes not only the average power level for each operating phase, but also the maximum power surge magnitude, load fluctuation duration, and average response delay, revealing the stability and dynamic trends of the operating condition in actual operation. For example, for a typical "hill start + hill cruise" operating condition, historical power records typically exhibit a sudden instantaneous power surge of 3-5 seconds during the start phase, while exhibiting a relatively stable output curve during the cruise phase. Therefore, the generated load profile reflects the dynamic power characteristics required by the system during different mission phases.

[0107] Secondly, the system divides the task objectives corresponding to the current working condition into time periods. This division is not only based on the running time, but also introduces the logical segmentation of the execution phase, such as "start-acceleration-constant speed-climbing-downhill", and sets the load level for each time period. The load level is quantified using a grading mechanism, for example, it is set to five levels from L1 to L5, corresponding to very low, low, medium, high, and very high load levels respectively. The system matches the load level of each time period with the load characteristic description extracted above, thereby giving each time period a preliminary power value estimate. Assuming that the first 3 seconds of the current task phase is the starting phase, its load level is calibrated as L5. The system will find the average power and maximum power values ​​of the corresponding phase L5 in the historical sample of the working condition, and assign a preliminary target power value to the current segment.

[0108] Next, in order to make the target power estimation closer to the actual operating status, the system also needs to introduce real-time environmental factors to correct the target power step by step. These environmental factors include but are not limited to the currently detected wind speed, temperature, humidity, terrain undulation, and obstacle density. By labeling the historical sample data with environmental factors and modeling the impact weights during the training phase, the system can establish a correction mapping relationship for the "impact of environmental changes on unit load power." For example, when the wind speed is greater than 8m / s and the vehicle is traveling against the wind, the system will increase the original power value by a correction of 10%-15%. Each environmental factor will dynamically adjust the target power value for the corresponding time period in units of time periods, so that the output power is adaptable.

[0109] Then, in order to further improve the controllability of the system's supply and demand matching, the system will jointly compare the target power demand sequence with the current battery SOC state. As an important medium for buffering and regulation, the battery's state of charge directly affects its ability to respond to sudden load changes. Therefore, the system first divides the battery state into three sub-states: "dischargeable", "rechargeable", and "maintain" according to the preset SOC range, and combines the power buffer bandwidth to determine whether it has the ability to access instantaneous surge loads or replace the generator output. If the conditions are met, the system will introduce a load transfer correction factor into the target power sequence, that is, introduce additional power in a certain period of time, which means that the battery will take on the load instead of the generator, or reversely absorb surplus power and charge during the load trough period, thereby achieving two-way coordination.

[0110] Finally, all adjusted target power value sequences are combined with the base loads of each phase to form a chronological set of total system power demand values. This result not only preserves the load characteristics of the current task, but also incorporates environmental influences and battery responsiveness. Furthermore, the model's adaptability to abnormal load fluctuations is enhanced by introducing historical interference immunity parameters. The resulting total system power demand value can be used for subsequent power scheduling of power generation units, battery charge and discharge management, and bus power allocation, achieving dynamic optimal control of the energy system under complex operating conditions.

[0111] This comprehensive power prediction method, which combines operating condition identification, historical sample statistics, real-time environmental perception, battery response modeling and phased adjustment factors, is a significant technological advancement. Different from existing power distribution schemes based on real-time load monitoring or simple empirical table scheduling, it has higher real-time performance, adaptability and power accuracy, and can effectively improve the overall energy supply efficiency and operational safety of the integrated power management system under alternating high and low load conditions.

[0112] Step S104: Determine the output power of the power generation unit and the battery charge and discharge power based on the battery state of charge (SOC) value and the total power demand value of the system. Under normal operating conditions, the output power of the power generation unit is determined to meet the basic load demand based on a preset SOC threshold range, and the remaining power is balanced by the battery for charge and discharge within the SOC range. Under emergency operating conditions, when the load suddenly increases, the battery is controlled to discharge at the maximum discharge rate. When the load suddenly decreases, the output power of the power generation unit is controlled to decrease rapidly, and the battery charging current is adjusted to avoid exceeding the maximum charging current that the battery can withstand.

[0113] In the comprehensive electric energy management and control method described in the present invention, the implementation of step S104 is crucial. Its core lies in scientifically and rationally determining the power distribution relationship between the power generation unit and the battery based on the total system power demand value calculated in the previous step and the current state of charge (SOC) value of the battery, thereby providing clear and executable output instructions for subsequent energy supply execution and system stability control.

[0114] In practice, this step first receives two key inputs: the current battery SOC value and the total system power demand value obtained in step S103. These two parameters represent the available capacity of the energy storage unit and the current real-time energy supply demand of the system. After receiving these inputs, the controller determines and allocates different strategy paths based on the current operating condition.

[0115] When the system is operating normally, indicating no sudden changes in load demand and the battery state is within normal limits, the control strategy prioritizes stable energy supply and efficiency. The power generation unit will serve as the primary energy source, and its output power should primarily cover the system's base load. To avoid wasted resources or excessive battery involvement, the controller prioritizes a preset SOC threshold range. This range is typically determined based on battery type, safety standards, and historical system operating experience; a range of 30% to 90% state of charge is generally recommended. If the SOC falls within this range, the system allows the battery to assist in providing power. Specifically, if the power generation unit's output slightly exceeds the current load, the battery enters a charging state to absorb excess energy. If the power generation unit cannot independently meet the load demand, the battery enters a discharging state to fill the shortfall. This entire process requires dynamic balance to ensure that while the battery participates in regulation, its SOC fluctuations remain within the permitted range, preventing overcharging and discharging.

[0116] When the system is judged to be in an emergency condition, the load behavior often changes dramatically, such as a surge in power demand in a short period of time, or the sudden disconnection of certain high-power loads, resulting in a reversal of supply and demand. The control strategy in such situations must have a fast response and protection mechanism. In the case of a sudden increase in load, the response speed of the power generation unit may not be enough to immediately follow the load increase, so the battery must be immediately controlled to discharge at the maximum allowable rate to quickly release electrical energy to support the supply and demand balance of the system. During implementation, the maximum rate should be determined by the safe discharge parameters specified by the battery manufacturer. The controller needs to monitor the battery temperature, current and internal voltage status in real time to ensure that the discharge process does not exceed the safety limit.

[0117] Conversely, in the event of a sudden load drop, the system faces the challenge of oversupply, potentially causing a transient increase in bus voltage. Prioritizing this, the power generation unit's output power should be rapidly reduced to its base operating power level while simultaneously switching the battery to charging mode to absorb the excess energy. To prevent thermal runaway or battery life degradation from high-current charging, it is necessary to further determine whether the current charging current is close to the maximum charge current the battery can withstand. The controller can dynamically limit the charging rate by referencing the SOC rise rate, the current sensor sampling value, and the internal protection threshold. If necessary, it can suspend power supply from the power generation unit until power returns to a stable state.

[0118] All of the above determinations and adjustments should be executed within millisecond cycles to ensure that the system's energy balance is not disrupted. It is recommended to introduce a state-triggered priority queue into the control strategy to prioritize responses to emergencies and avoid uncertain behavior caused by conflicting control commands. To ensure safety, a protection priority strategy can also be set. When the discharge rate or charging current is about to exceed the limit, the controller immediately suspends all other optimization controls and forcibly switches into protection logic to ensure the safe operation of the battery and power generation unit.

[0119] Therefore, the technical core of step S104 is not only to simply allocate power, but also to achieve smooth, precise and controllable power supply strategy under normal and emergency working conditions based on multiple judgments and coordinated control of dynamic changes in supply and demand, energy storage response capability and safe operation boundary, to ensure that the entire power system can maintain efficient, stable and safe operation state under various complex load scenarios.

[0120] Further, the determination of the output power of the power generation unit and the charging and discharging power of the battery according to the battery state of charge SOC value and the total power demand value of the system comprises:

[0121] According to the historical basic load model corresponding to the working condition label and the current working condition mode information, the target power demand value of the system at the current time is extracted, and the basic power part in the demand value is preliminarily set as the power output target of the power generation unit;

[0122] The current SOC value of the battery is read, and the SOC threshold interval is compared to determine whether it is in the charging area, discharging area or balance maintenance area, and the acceptable battery charging and discharging power adjustment range is formed accordingly;

[0123] If the current working condition belongs to normal working condition, under the premise of meeting the basic load demand, according to the interval position of the SOC value and the load change trend, the remaining power is allocated for corresponding charging or discharging operation by the battery, and the joint allocation result of the output power of the power generation unit and the charging and discharging power of the battery is generated;

[0124] If the current working condition belongs to emergency working condition, and the power demand presents a sudden increase trend, the discharging power of the battery is set as the maximum value allowed at the current time under the constraint of not exceeding the maximum rate discharging capability of the battery, and the power of the power generation unit is temporarily maintained or slightly increased to stabilize the bus voltage;

[0125] If it is detected that the load appears rapid sudden reduction, the output power of the power generation unit is quickly reduced, and the battery charging current limit model is called to set the charging power value of the battery under the premise of not exceeding the maximum safe charging current of the battery, and finally the target output power value of the power generation unit at the current time and the charging and discharging power value of the battery are output.

[0126] After the system receives the current working condition mode information and the target power demand value, first, the standard basic power value under the working condition is extracted based on the historical basic load model corresponding to the working condition label. The historical basic load model can be constructed by long-term recording of the power consumption curves of the vehicle or system under different working conditions, especially by extracting the load fluctuation data under typical operating states such as cruising, climbing, vertical take-off and landing, and silent driving, and statistically clustering them to form a standardized load reference set. In this process, to avoid error accumulation, each working condition label needs to match the time window corresponding to its typical operating section to ensure that the obtained load baseline is representative and stable.

[0127] The system further dynamically corrects the basic power in combination with the path planning of the current working condition and real-time weather interference factors (such as wind speed, electromagnetic disturbance, etc.) to form the target power demand value at the current time of the system. The target value includes static demand (basic load) and dynamic correction amount, and the dynamic correction amount is derived from the disturbance of environmental factors on the power consumption behavior of the system, such as the propulsion power fluctuation caused by wind resistance or the influence of external interference on the cooling load of the system.

[0128] Subsequently, the system reads the state of charge (SOC) value of the current battery, dynamically corrects the SOC value by combining the real-time information of voltage, current, temperature, etc. provided by the battery management system, and the embedded temperature compensation model. The temperature compensation model considers the nonlinear change of battery discharge efficiency at different temperatures, for example, introduces a temperature coefficient κ(T) = 0.96 at 25°C, and adjusts the effective value of current used in the current integration formula under low or high temperature conditions to obtain more accurate SOC estimation. The SOC interval is pre-set to three ranges: charging zone, discharging zone and balance maintenance zone. Generally, SOC below 20% is considered as the discharging zone, and above 80% is considered as the charging zone, and the middle is the balance maintenance zone. The specific threshold can be set according to the battery characteristics in the product design stage and fixed in the control strategy.

[0129] If the system identifies the current working condition as a normal working condition (such as cruising flight or land charging), the control strategy takes the basic load as the core, and judges whether the remaining power can be used for battery charging or whether the battery needs to be discharged to supplement the short-term fluctuation based on the load. Specifically, if the current SOC value is at the upper limit of the charging interval, it means that the battery is close to full charge, at which time the system avoids forced charging and selects the power generation unit to independently bear all the load; if it is at the lower limit of the discharging interval, it means that the battery is insufficient, at which time the power generation unit is used to bear most of the load as much as possible to avoid the battery from continuing to discharge. In the balance maintenance interval, the system sets the output power of the power generation unit to stably cover the basic load based on the short-term trend prediction result of the load, and the battery provides the rapidly changing part of the power to balance the load fluctuation and improve the response sensitivity of the system.

[0130] If the current operating condition is identified as an emergency, the control strategy will quickly enter the dynamic adjustment phase. In the event of a sudden load surge, the system first retrieves the battery's maximum discharge rate capability. This rate is fixed to a multiple of the rated capacity (such as 2C or 3C) based on the battery's design parameters. Once this limit is confirmed, the control system sets the battery discharge power to the current maximum allowable value to quickly respond to the load surge and ensure system voltage stability. At the same time, the current output power of the power generation unit is maintained unchanged or slightly increased, and the bus voltage is supported by two parallel power supply channels.

[0131] If a sudden load drop occurs, that is, the system load demand drops significantly, the system needs to avoid the risk of bus overvoltage caused by continuous high power output of the power generation unit. Therefore, the target output power of the power generation unit is quickly lowered first, and the built-in battery charging current limit model is called to obtain the maximum safe charging current value under the current temperature and SOC conditions. The model determines the current maximum charging power by looking up the table or interpolation based on the battery charging characteristic diagram. For example, under the conditions of SOC of 62% and temperature of 25°C, the maximum safe charging current is 6A and the voltage is 51.2V, then the maximum charging power is approximately 307W. The system uses this power as the upper limit to control the battery to enter the current limiting charging mode to ensure its safety and life are not affected.

[0132] Ultimately, based on the judgment results of each path, the system summarizes the target output power of the power generation unit and the target charge and discharge power of the battery to form a combined power output configuration plan for the energy supply unit. This plan is used to drive the actual energy distribution behavior in the next control cycle, ensuring efficient and stable operation of the entire system under complex operating conditions and load disturbances.

[0133] For example, in a hybrid energy supply system, the base load under normal cruising conditions is set to a constant 5.0 kW. The energy supply system consists of a diesel generator unit with a rated power of 8.0 kW and a lithium battery pack with a rated capacity of 10 Ah and a nominal voltage of 51.2 V. The system is currently executing a path planning operation with the operating mode information set to "low-speed climbing" and a target power demand of 6.2 kW. At this time, the battery SOC obtained by the system status monitoring module is 62%, the ambient temperature is 25°C, the real-time current is 4.5 A, and the bus voltage is 51.2 V.

[0134] Firstly, the system identifies the current operating condition as "low-speed climbing", and establishes a matching relationship with the historical operating condition database through the operating condition label. The database is constructed by historical operation data collection, and the structure is in the form of <operating condition label, time period, average basic load, standard deviation>. Under the label of "low-speed climbing", the average basic load is 5.0 kW, and the load standard deviation is 0.4 kW. The system combines the slope information (2.8%) and wind speed information (gentle breeze, wind resistance correction coefficient is 1.05) in the current path, and according to the environmental intervention correction rule, the basic load is corrected to 5.0 × 1.05 ≈ 5.25 kW.

[0135] The system then calls the battery SOC evaluation module to make corrections according to the current readings. The initial SOC value is 62%, and the temperature is 25°C. According to the temperature compensation curve constructed by experience, the current temperature compensation coefficient κ(T) = 0.96. Since the system uses current integration to calculate SOC, specifically, the SOC at the previous time is subtracted by the ratio of discharge capacity to battery rated capacity in unit time, but due to the high volatility of current, the system uses small time window integration to make real-time corrections. In this example, the current is stable at 4.5 A per unit time, so according to the estimation, the power change in one minute is about 4.5 A × 1 min = 0.075 Ah, which is 0.75% of the total capacity of 10 Ah. After temperature compensation, the actual discharge effect is 0.75% × 0.96 = 0.72%. Therefore, the system updates the SOC to about 61.28%.

[0136] Next, according to the preset SOC threshold interval to determine the charging and discharging strategy, set: SOC < 30% as low area and need to charge first, SOC > 80% as high area and need to limit charging, 30% ~ 80% as adjustable area. The current SOC is 61.28%, which belongs to the adjustable area. Then the system enters the load distribution strategy judgment process.

[0137] Because the current total power demand is 6.2 kW, the basic load is 5.25 kW, and the remaining 0.95 kW is adjusted by the battery. At this time, the system judges that the 0.95 kW is compensation power. Since the SOC is in the adjustable area, the system further combines the load short-term prediction model to determine the direction. The prediction model is constructed by simple moving average method, and the power change trend in the last 5 minutes is calculated. In the last 5 minutes, the power is 5.4, 5.6, 5.8, 6.0, 6.2 kW, showing a linear growth trend, so the next period will still increase. The system therefore considers this trend as the need to increase output, so it decides that the battery will perform auxiliary discharge operation.

[0138] In the emergency judgment branch, assume a sudden load increase occurs during the current cycle. The system records a power change exceeding 1.2 kW per unit time, deeming it a sudden increase and initiating emergency operation. Based on the battery specifications, the maximum discharge rate is set at 2C, meaning the maximum discharge current is 20 A (2 × 10 Ah) and the maximum discharge power is 20 A × 51.2 V = 1.024 kW. The current load increase is 1.2 kW, exceeding the maximum discharge capacity. Therefore, the system sets the battery discharge power to 1.024 kW, with the power generation unit increasing its output by 0.176 kW to support the transient load.

[0139] In a sudden load drop scenario, assuming a sudden drop from 6.2 kW to 4.2 kW, the system determines a 2 kW drop. The emergency strategy quickly reduces the power generation unit to 4.2 kW and introduces a battery charging current limiter. This model predefines a maximum safe charging current of 8 A (corresponding to the battery's internal thermal response model), resulting in a maximum charging power of 8 A × 51.2 V = 409.6 W. The system controls the battery to charge at only 409.6 W, preventing the generation unit from outputting excess power to prevent energy recirculation or overvoltage risks.

[0140] Finally, based on the results of real-time dynamic judgment, the system outputs the energy supply configuration plan for the current cycle: the output power of the power generation unit is 4.2 kW, and the battery charging power is 409.6 W, ensuring the stable operation of the bus while taking into account battery safety and efficient energy utilization.

[0141] Step S105: controlling the actual energy supply behavior of the power generation unit and the battery according to the output power of the power generation unit and the charge and discharge power of the battery.

[0142] In the comprehensive electric energy management and control method described in this invention, step S105 aims to effectively convert the power generation unit output power and battery charge / discharge power, determined in the previous step, into specific control instructions, driving the system's actual energy supply components to respond accurately, thereby completing the closed-loop transition from theoretical energy allocation to actual power supply behavior. The technical implementation of this step involves not only the generation and distribution of power control signals but also the real-time coordination and status feedback mechanism between the power generation unit and the battery management system. Its correct execution is crucial for stable system operation.

[0143] In actual execution process, the controller first calls the power distribution data output in step S104, which contains the target output power value that the power generation unit should currently bear, and the charging or discharging power range that the battery should execute under the current working condition. The system needs to judge the main energy supply path under the current running mode based on the value, that is, whether to be supplied by the power generation unit or the battery, or both. At this time, the controller needs to generate matching control instructions according to the current working state of each energy supply channel, and issue the instructions to the power generation control module and the battery management module respectively.

[0144] In terms of the power generation unit, if the system structure is an internal combustion generator matched with an electric control inverter module, the controller needs to adjust the throttle opening or excitation control signal according to the power target value, so that the generator keeps synchronous with the load demand in terms of output frequency, voltage and current, etc. At the same time, if the generator outputs power to the system through a DC bus, it also needs to control the output channel parameters of the inverter or DC-DC converter to ensure voltage matching and current limit controlled operation in the energy transmission process. In addition, when the generator power changes rapidly, current slope limiting logic should be supplemented to prevent system shock caused by sudden output.

[0145] In terms of battery energy supply, the controller transmits the target charging or discharging power to the battery management system (BMS). The BMS controls the on-off state of the battery according to the instruction, and combines its internal voltage, current, temperature and SOC parameters to judge whether it has corresponding discharging or charging capacity. If the conditions are met, the BMS controls the on-off and current size of the output interface to make the battery perform the corresponding energy exchange behavior according to the controller's requirements. If the current SOC is close to the upper and lower thresholds or the battery has the risk of over temperature, the BMS has the right to feedback the "refuse to respond" signal, requiring the controller to adjust the allocation scheme again and enter the degraded operation or switch to the power generation unit main supply mode. Such feedback mechanism should be set as a high-speed response channel to ensure that the system completes the path switching within milliseconds to avoid power interruption.

[0146] In the joint working scenario of power generation unit and battery participating in energy supply at the same time, the controller should introduce power synchronization and coordination logic. This logic compares the current, voltage and load coupling relationship of the generator output end and the battery output end in synchronization, dynamically adjusts the power ratio between them, and adjusts the output of both sides through the voltage fluctuation trend of the load end bus to avoid mutual interference or short-time reverse current, etc. In addition, in this joint energy supply mode, it is recommended to use a master-slave control architecture, in which the power generation unit provides the reference power curve and the battery provides the dynamic compensation margin to achieve higher response speed and smaller bus voltage fluctuation range.

[0147] To ensure the accuracy of the control results, the controller should receive the execution feedback information of the power generation module and the battery module in real time after the control command is issued, and continuously monitor the key parameters such as output power, output voltage and working state. If the correct feedback is not received within the specified time, or the output power deviates from the set target and exceeds the warning range, the system should immediately enter the abnormal processing procedure, including power redistribution, standby power generation unit switching in, battery charging and discharging mode reversal, or temporary load reduction mechanism, so as to ensure uninterrupted power supply and suppress bus voltage fluctuation.

[0148] In summary, the implementation of step S105 not only involves basic signal transmission and device response process, but also is the execution guarantee mechanism of the overall power management strategy of the control system, which requires the controller to have high-precision control algorithm, fast response logic, interface adaptation ability with energy supply hardware devices, and complete safety protection and abnormal processing system. Only on this basis, the two types of energy modules, power generation unit and battery, can realize multiple control goals such as cooperative energy supply, dynamic adjustment, load matching and system protection under normal and emergency conditions.

[0149] Further, the control of the actual energy supply behavior of the power generation unit and the battery according to the power generation unit output power and the battery charging and discharging power includes:

[0150] Receiving the determined power generation unit output power and battery charging and discharging power, and dynamically comparing the sum of the two with the target bus power demand value to identify whether the current total energy supply meets the system instantaneous power balance requirement, and generating an energy supply balance adjustment signal when detecting deficiency or redundancy;

[0151] According to the energy supply balance adjustment signal, the dynamic adjustment limit of the power generation unit under the current mechanical speed, fuel supply and environmental temperature conditions is calculated first, a set of power generation power adjustment boundary interval is formed, and an output is used for the next power adjustment judgment;

[0152] Based on the current SOC value of the battery, its temperature, aging degree and current cycle rate, the maximum charging current and maximum discharge rate parameters are queried from the preset parameters, a battery adjustment capability description vector is generated, and a joint comparison is made with the above-mentioned power generation power adjustment boundary interval to deduce a multi-strategy energy supply combination that can be adjusted in this round;

[0153] According to the current load type, the predicted power fluctuation trend and the energy utilization priority rules, the optimal strategy is selected from all feasible energy supply combinations. If the predicted short-time load fluctuation is frequent, the battery is preferred; if the load is stable for a long time, the power generation unit is preferred to bear the basic energy supply, so as to determine the target execution output value of the power generation unit and the battery in each control period;

[0154] The target execution output values of the power generation unit and the battery are respectively converted into corresponding control instruction parameters, including speed setting, excitation adjustment, PWM control, and current loop control, and are respectively issued to the power generation unit and the battery management system through an execution control module to drive the actual energy supply operation and realize dynamic energy collaborative supply based on power demand closed-loop control.

[0155] In the power comprehensive management control method, for the actual energy supply behavior control process of the power generation unit and the battery, the target output power of the power generation unit and the target charge-discharge power of the battery determined by the system according to the target power demand, the working condition category, the battery SOC state and other multi-dimensional information in the previous control period are first received. These two power values constitute the total energy supply amount preliminarily planned by the system in the current control period. Before the actual energy supply control is executed, the total energy supply amount and the target bus power demand value on the actual load side need to be compared and verified. The system obtains the instantaneous power demand value of the current system bus through a high-speed sampling circuit and a bus real-time power monitoring module, and performs difference operation on the sum of the power generation unit and the battery power. If the difference is within the allowable range, it means that the energy supply plan can meet the current load balancing demand and no further adjustment is needed. If the difference deviates from the normal range, the system will immediately generate an energy supply balancing adjustment signal to indicate that there is a risk of power shortage or redundancy, which must be adjusted through the power generation unit and the battery to maintain the stability of the bus power.

[0156] After receiving the energy supply balancing adjustment signal, the system needs to determine whether the power generation unit has adjustment space. For this purpose, the current mechanical speed, fuel injection pulse width, air intake temperature and external environment temperature of the power generation unit are read in real time, and these parameters are input into the power generation unit dynamic capability model for boundary prediction. The model is constructed based on the engine dynamic response curve and the load dynamic characteristic database, and can output the maximum power-up capability and the maximum power-down capability boundary of the power generation unit under the current conditions. For example, in the scenario where the fuel injection pulse width is unchanged and the external temperature rises, the maximum power-up capability of the power generation unit will be limited due to the decrease of cooling capacity, so the model will timely shrink the upper limit of adjustment, so as to ensure that the power generation unit operates within the safe range. The output result of the boundary model is expressed in the form of upper limit and lower limit, forming the boundary interval of the adjustable power of the power generation unit in this adjustment period.

[0157] In parallel with the power generation unit capacity boundary, the system also needs to evaluate the dynamic adjustment capability of the battery. To this end, the system will read the current SOC value of the battery, and in combination with real-time temperature information, battery aging assessment indicators (such as number of cycles, internal resistance change rate), and current rate discharge status, query the preset battery safety operating parameter table. This parameter table lists the maximum charging current, maximum discharge current and its duration corresponding to different SOC intervals at different temperatures and aging levels, and is corrected with a safe operation factor. Based on this, the system will construct a set of battery adjustable power range vectors, including key parameters such as the current maximum allowable discharge power, maximum charging power, and optimal working rate, and then jointly compare this vector with the aforementioned power boundary interval of the power generation unit. This comparison process is used to identify all possible energy supply combinations that can maintain supply and demand balance and do not exceed the adjustment capability of any energy supply component during this round of adjustment cycle.

[0158] After obtaining multiple feasible energy supply combinations, the system needs to optimize them based on the current load type and its power fluctuation trend. For example, when the prediction model determines that the load will show a trend of frequent and high-amplitude fluctuations in the next few control cycles (such as the start and stop of welding equipment, the activation of impact loads, etc.), a battery-based fast response strategy is preferred, so that the battery takes on the power transient regulation, while the power generation unit maintains stable operation to avoid mechanical loss caused by frequent load increases and decreases. If the load trend is predicted to be long-term stable, the power generation unit is preferentially dispatched to take on the basic power output, and the battery only participates in peak and valley regulation or energy recovery operations. The judgment of the load fluctuation trend is based on historical load data, real-time fluctuation amplitude indicators and prediction model output. The prediction model can be established based on technologies such as sliding window regression analysis and power spectrum density analysis, and has a high instantaneous judgment accuracy.

[0159] After selecting the optimal energy supply strategy, the system will assign target execution output values ​​for the power generation unit and battery within the current control cycle. For the power generation unit, this output value will be converted into specific control instructions such as the engine speed setpoint, fuel injection amount setpoint, cooling fan speed control value, and excitation current adjustment value. These control quantities will be sent to the power generation unit controller via the CAN bus or Modbus communication protocol in a standardized command format to achieve refined output control. For the battery, the target execution output value will be converted into control signals such as current command, voltage setting, current limit flag, and battery protection logic control switch status, and sent to the battery management system (BMS). The BMS will execute current loop control, voltage loop regulation, and overcurrent / overvoltage / undervoltage protection logic to ensure safe and efficient charging and discharging.

[0160] During the energy supply control execution process, the system also establishes a fast feedback channel. Through high-frequency sampling feedback of actual output power of the power generation unit, actual charging and discharging current of the battery, bus voltage, power error and other data, the feedback is fed back to the energy control center for the next control period to adjust the prediction model and optimization strategy parameters, so as to realize closed-loop control. The whole process is rolling forward in units of control periods. Each round of execution is based on ensuring the dynamic balance of power supply and demand, and taking into account component safety, energy efficiency and response timeliness, so as to realize the comprehensive management strategy of electric energy based on high and low load condition identification and demand response.

[0161] Step S106: Based on the actual energy supply behavior, the system bus voltage and the battery state of charge SOC value are detected. If the bus voltage or the battery SOC value deviates from the set range, the output power of the power generation unit and the charging and discharging power of the battery are adjusted according to the deviation, until the bus voltage and the battery SOC value return to the allowed range.

[0162] In the comprehensive management control method of electric energy of the application, the core purpose of step S106 is to establish a dynamic closed-loop adjustment mechanism based on energy supply behavior feedback, to ensure that the system bus voltage and the battery state of charge (SOC) always operate stably within the preset safe range. Due to the comprehensive interference of various factors such as load fluctuation, battery state change, environmental temperature influence and other factors during the operation of the energy supply system, the bus voltage and the battery SOC may deviate from the expected value in a short time. If not adjusted in time, it may cause energy management disorder, battery wear and tear, and even system energy supply interruption. Therefore, this step provides a strategy for real-time detection and feedback control cooperation, so that the system always maintains the stability and safety of electrical parameters during operation.

[0163] Specifically, after the energy supply control instruction is issued and executed, the system collects the bus voltage in real time through the voltage sensor, and monitors the current SOC value by the battery management system, and feeds the two key operating states to the central control module. Among them, the bus voltage reflects whether the total power between the current power generation unit and the battery output meets the load demand. If the voltage rises too fast, it may mean that the battery charging current is too large or the load is too light. If the voltage drops sharply, it may be due to sudden load increase or power supply response lag. The SOC value is a core indicator to measure the battery energy storage status, and its trend directly determines the upper and lower limits of the subsequent charging and discharging capacity allocation.

[0164] After obtaining the latest bus voltage and SOC value, the control module first compares it with the normal operating range preset in the system. This range is usually determined by engineering experience and system characteristics. For example, the bus voltage can be set to a range with a ±5% fluctuation, and the battery SOC range is generally 20% to 90% to prevent battery life decline or safety hazards caused by overcharging or overdischarging.

[0165] If the monitoring value is within the set range, the system maintains the original control strategy unchanged, and only continues to monitor as a warning; and once any indicator is found to be out of the range, the closed-loop adjustment logic is triggered. The adjustment process takes the deviation size as input to determine whether the current deviation trend is upward or downward, and accordingly formulates a response strategy. If the bus voltage is too high and the battery SOC is also close to the upper limit, the system will preferentially reduce the output power of the power generation unit and limit the battery charging current; if the bus voltage is too low and the SOC is close to the lower limit, the output of the power generation unit should be increased or the standby power supply should be started, while the battery discharge rate is controlled to reduce, and if necessary, even stop the discharge behavior to prevent over-discharge.

[0166] In the closed-loop adjustment strategy, the system adopts a feedback mechanism with proportional gain, that is, the adjustment strength is determined according to the deviation amplitude. For example, if the bus voltage deviates slightly from the allowed range, the controller can only fine-tune the output voltage or frequency of the power generation unit to quickly restore stability; if the deviation is serious, multi-parameter coordinated adjustment is started, including changing the battery current limit, switching the load priority, and other multi-level measures to achieve more rapid and stable regression.

[0167] In addition, in order to improve the robustness and adjustment accuracy of the system, it is suggested that the control process introduce a hysteresis judgment and a dithering elimination mechanism to avoid frequent triggering of adjustment instructions due to sampling fluctuations, thereby causing system oscillation. In specific implementation, a minimum deviation threshold can be set, and only when the deviation exceeding the threshold lasts for more than a certain period of time, the adjustment is started to ensure that the system still has fault tolerance and buffering capability under slight disturbance.

[0168] After the bus voltage and battery SOC are restored to the allowed range, the controller should automatically mark the system running state as "stable" and re-initialize the monitoring and adjustment logic of this round to enter the next cycle of the closed-loop monitoring process of the power supply behavior. The whole step not only realizes the dynamic tracking and rapid response of the system power supply state, but also strengthens the safety guarantee mechanism in energy flow control, which is a key component of the invention to realize adaptive control under high and low load conditions. This technical path is suitable for new energy vehicle power supply, distributed micro-grid power supply system and mobile energy platform and other types of scenes, and has strong universality and practicality.

[0169] Further, based on the actual power supply behavior, the system bus voltage and battery state of charge (SOC) value are detected, and if the bus voltage or battery SOC value deviates from the set range, the output power of the power generation unit and the battery charging and discharging power are adjusted according to the deviation in a closed-loop manner until the bus voltage and battery SOC value are restored to the allowed range, including:

[0170] In each control cycle, the voltage value of the system bus and the SOC value of the battery are obtained in real time, and are compared with the respective dynamic set threshold interval to determine whether there is a deviation beyond the range, wherein the bus voltage threshold is dynamically set based on the current load type, power grid structure and target stability margin, and the battery SOC threshold is adjusted in combination with the current temperature, discharge rate and cycle aging degree;

[0171] If it is found that any indicator deviates, a regulation sequence is triggered, first by consulting the preset deviation response mapping table to obtain the recommended priority regulation resource and regulation direction corresponding to the deviation amplitude, and combining the power dynamic data in the current control cycle to construct a preliminary response strategy plan;

[0172] According to the preliminary response strategy plan, the actual power interval of the power generation unit that can be adjusted is calculated, and the current acceptable charge and discharge rate limit of the battery is synchronously evaluated to form the upper limit of the power generation unit adjustment and the battery regulation capability interval, respectively, and the two groups of boundary parameters are taken as inputs to generate multiple groups of combined regulation configurations for selection;

[0173] Among the multiple groups of combined configurations, the system selects the regulation scheme that does not cause new load disturbance, thermal management exceeding the standard or battery life loss intensifying according to the energy utilization priority ordering rule, finally determines an optimal regulation path, and accordingly issues regulation instruction parameters to the power generation unit and the battery respectively;

[0174] After the regulation path is executed, a high-speed feedback channel is established to continuously monitor the recovery progress of the bus voltage and the battery SOC value, and dynamically adjust the regulation amplitude and the control cycle rhythm until both are stable and return to the allowed range, and the regulation process is terminated and enters the next regular control cycle.

[0175] In high and low load conditions, to realize fine regulation and control of the operation state of the electric energy system, the present application proposes a voltage and state of charge double closed loop regulation mechanism based on actual energy supply behavior feedback, which is used to dynamically adjust the power generation unit output power and the battery charge and discharge power when the system bus voltage or battery state of charge SOC value deviates from the allowed operation interval, so as to realize the stable regression of the bus voltage and the battery SOC value, and guarantee the overall operation safety and energy utilization efficiency of the system.

[0176] First, during each control cycle, the system uses a monitoring module to collect the current bus voltage and battery state-of-charge (SOC) values. The real-time bus voltage is typically collected via high-precision voltage sensors installed at both ends of the bus, while the SOC value is calculated by the battery management system (BMS) using a fusion method combining current integration, voltage estimation, and a historical calibration model. To ensure dynamic adaptability, the system dynamically generates reasonable threshold ranges for the bus voltage and battery SOC before performing a comparative judgment based on the current operating load type, grid structure stability requirements, battery temperature, cycle count, and other status information. For example, if the current load type is a motor-driven load, the allowable voltage fluctuation range should be appropriately expanded to buffer transient shocks. Under low-temperature conditions, the battery's charge and discharge capacity decreases, so the lower SOC limit should be appropriately increased to avoid the risk of over-discharge.

[0177] After completing the comparison of the voltage and SOC values ​​with the threshold range, if any indicator is found to have deviated, the activation of the adjustment sequence will be triggered immediately. After the adjustment sequence is started, the system will first call the preset deviation response mapping table, which records the recommended adjustment resource allocation scheme, response strategy priority and power adjustment direction corresponding to different deviation types and amplitudes. The construction of the deviation response mapping table is based on long-term operating data and expert experience rules, and can be updated according to the online operation status of the system. The system combines the power dynamic behavior recorded in the current control cycle, including the load power change trend, the current energy supply combination mode, energy utilization efficiency, etc., to generate a set of preliminary response strategy plans. The plan will clearly indicate whether to prioritize the adjustment of the power generation unit or battery output power, the initial value of the adjustment amplitude and the expected adjustment duration.

[0178] Next, based on the above preliminary strategic plan, the system calculates the adjustable capacity ranges of the power generation unit and the battery respectively. For the power generation unit, its maximum output power limit is affected by the current mechanical speed, fuel supply rate and cooling system capacity. The system reads these operating data in real time and calculates the upper and lower limits of the power generation power in combination with the equipment performance curve model. For example, when the ambient temperature is high, the maximum power output limit of the power generation unit will automatically be reduced to avoid overheating. At the same time, the battery's charge and discharge capacity will also be comprehensively evaluated based on multiple indicators such as the current SOC value, temperature status, current rate and historical cycle count. The system combines the parameter table provided by the battery manufacturer with its own charge and discharge capacity calculation model to determine the currently acceptable upper limit of the charging current and the discharge rate boundary, forming a description vector of the battery's adjustment capacity.

[0179] The system then compares the power generation unit's adjustment interval with the battery's regulation capability vector, deriving a set of multi-strategy regulation configuration combinations based on this. Each combination contains a set of possible power generation unit output power and battery charge and discharge power allocation schemes that meet their respective boundary constraints. To select the optimal strategy among these schemes, the system introduces an energy utilization priority sorting rule, which dynamically assigns weights based on system objectives (such as maximum efficiency, minimum loss, and minimum disturbance). For example, when power demand is predicted to fluctuate frequently in the short term, the system will give priority to batteries to take on the main regulation tasks to reduce the mechanical losses caused by the frequent speed changes of the power generation unit; when the power load is expected to operate stably for a long time, the system will give priority to using the power generation unit to provide basic power to extend the battery life and reduce its deep cycle number.

[0180] After the optimal regulation strategy is selected, the system converts the target power of the power generation unit and the target battery charge and discharge power defined in the strategy into specific control command parameters, which are then sent to the power generation control unit and battery management system via control interfaces. For the power generation unit, these control commands typically include parameters such as the excitation current setpoint, fuel injection rate, fan speed, and generator voltage setpoint; for the battery, these include information such as the target charge and discharge current, maximum allowable voltage, and current regulation slope. These parameters are then applied to the specific execution units through PWM control, voltage and current loop regulation, and analog-to-digital conversion interfaces.

[0181] After the adjustment instruction is issued, the system will not stop its monitoring behavior, but will continue to monitor the changes in bus voltage and SOC value through the high-speed feedback channel, and compare it with the set recovery target, and continuously evaluate whether the adjustment effect meets the expectations during the adjustment process. If it is found that the recovery trend is not obvious or the adjustment process causes new offsets in other parameters, the system will immediately trigger the next round of adjustment path calculation, update the adjustment amplitude or adjust the target value. For example, when the power generation unit has reached the output limit and the bus voltage has not returned, the system will try to further compress the battery discharge limit or temporarily reduce some non-critical loads. The entire closed-loop adjustment will be officially terminated after the bus voltage and SOC value return to the allowable range, and the system status returns to the normal scheduling cycle.

[0182] The regulation method of the present invention also possesses a certain degree of self-learning capability. In actual operation, the system records the response effect of each regulation sequence, analyzes the correlation between the regulation scheme and the regulation results, and gradually optimizes the deviation response mapping table and energy priority sorting rules to form an adaptive regulation knowledge base for different operating conditions. Through this mechanism, the system can gradually improve response speed, reduce the risk of over-regulation, and enhance its adaptability to complex operating conditions.

[0183] A second embodiment of the present application provides an electronic device, comprising:

[0184] processor;

[0185] The memory is used to store a program. When the program is read and executed by the processor, it executes an electric energy comprehensive management and control method for high and low load conditions provided in the first embodiment of the present application.

[0186] The third embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the program executes an electric energy comprehensive management and control method for high and low load conditions provided in the first embodiment of the present application.

[0187] Although the present application is disclosed as above with the preferred embodiments, it is not intended to limit the present application. Any person skilled in the art may make possible changes and modifications without departing from the spirit and scope of the present application. Therefore, the scope of protection of the present application shall be based on the scope defined by the claims of the present application.

Claims

1. A comprehensive power management and control method for high and low load conditions, characterized in that: include: Obtain the current operating information of the energy supply system, including operating mode information, target power demand, and current battery state of charge (SOC) value; Determining the current operating condition category based on the current operating information; Calculate the corresponding total system power demand value according to the operating condition category and the target power demand; The output power of the power generation unit and the battery charge and discharge power are determined based on the battery state of charge (SOC) value and the total power demand of the system. In normal operating conditions, the output power of the power generation unit is determined to meet the basic load demand based on a preset SOC threshold range, and the remaining power is balanced by the battery charge and discharge within the SOC range. In emergency operating conditions, when the load suddenly increases, the battery is controlled to discharge at the maximum discharge rate. When the load suddenly decreases, the output power of the power generation unit is controlled to decrease rapidly, and the battery charging current is adjusted to avoid exceeding the maximum charging current that the battery can withstand. Controlling the actual energy supply behavior of the power generation unit and the battery according to the output power of the power generation unit and the charge and discharge power of the battery; Based on the actual energy supply behavior, the system bus voltage and the battery state of charge (SOC) value are detected. If the bus voltage or the battery SOC value deviates from the set range, the power generation unit output power and the battery charge and discharge power are adjusted in a closed-loop manner according to the deviation until the bus voltage and the battery SOC value return to the allowable range; The method includes detecting the system bus voltage and the battery state of charge (SOC) value based on the actual energy supply behavior. If the bus voltage or the battery SOC value deviates from a set range, the power generation unit output power and the battery charge and discharge power are adjusted in a closed loop according to the deviation until the bus voltage and the battery SOC value return to an allowable range. The method includes: During each control cycle, the system bus voltage and battery SOC values ​​are acquired in real time and compared with their respective dynamically set thresholds to determine whether there are any deviations beyond the range. The bus voltage threshold is dynamically set based on the current load type, grid structure, and target stability margin, while the battery SOC threshold is adjusted based on the current temperature, discharge rate, and cycle aging. If any indicator is found to have a deviation, the adjustment sequence is triggered. First, the preset deviation response mapping table is consulted to obtain the recommended priority adjustment resources and adjustment direction corresponding to the deviation amplitude. Combined with the power dynamic data in the current control cycle, the preliminary response strategy plan required for this round of adjustment is constructed. Based on the preliminary response strategy plan, the actual adjustable power range of the power generation unit is calculated, and the current acceptable charge and discharge rate limit of the battery is simultaneously evaluated. These two sets of boundary parameters are used as input to generate multiple sets of combined adjustment configurations for selection. Among multiple configuration combinations, the system prioritizes energy utilization and prioritizes those that do not cause additional load disturbances, exceed thermal management standards, or increase battery lifespan loss. Ultimately, it determines a set of optimal regulation paths and issues adjustment command parameters to the power generation units and batteries accordingly. After the regulation path is executed, a high-speed feedback channel is established to continuously monitor the recovery progress of the bus voltage and battery SOC value, and dynamically adjust the regulation amplitude and control cycle rhythm until both are stabilized and return to the allowable range. The regulation process is terminated and enters the next regular control cycle.

2. The electric energy comprehensive management and control method for high and low load conditions according to claim 1 is characterized in that: The operating condition categories include normal operating conditions and emergency operating conditions, wherein the normal operating conditions include parking charging, silent driving, land charging, vertical take-off and landing, and cruise flight, and the emergency operating conditions include emergency braking, high-power maneuvering, fan damage, and emergency obstacle crossing.

3. The electric energy comprehensive management and control method for high and low load conditions according to claim 1 is characterized in that: The obtaining of the current operating information of the energy supply system includes: Based on the task instructions or operation control instructions received by the energy supply system, the system analyzes the implicit operation phase, driving status, and environmental adaptation strategy, identifies the corresponding operating condition labels, and extracts and generates the current operating mode information; After obtaining the working condition mode information, the reference power data corresponding to the working condition tag in the historical load model and environmental database is retrieved, and combined with the real-time measured load current and voltage change trends, the target power demand required at the current moment is deduced and output; After the target power demand is determined, the battery voltage, current and temperature data fed back by the battery management system are read in real time. The remaining battery capacity is estimated using the Coulomb integration method corrected by temperature compensation, and the current battery state of charge (SOC) value is calculated and output. Based on the current target power demand and SOC value, determine whether the battery can meet the energy supply demand without exceeding the safe operating boundary, and compare the judgment result with the aforementioned operating mode information to confirm whether there is a risk of power supply and demand mismatch.

4. The electric energy comprehensive management and control method for high and low load conditions according to claim 1 is characterized in that: The determining of the current operating condition category based on the current operating information includes: Based on the operation target semantic fields, path planning labels and dynamic environment intervention items extracted from the operation mode information, a task context semantic vector is constructed and mapped to a multidimensional operation condition label embedding space; In the multi-dimensional operating condition label embedding space, a dynamic clustering algorithm is used to identify the closest historical operating condition cluster center, and combined with the current target power demand change trend, a preliminary operating condition classification result and a corresponding credibility score are output; Based on the preliminary operating condition classification results, the dynamic energy consumption curve family and the charge safety limit parameter group associated therewith are called to construct a safety mapping space of the battery state of charge (SOC) value, and the current SOC value position is combined with the determination of whether it is in the controllable load window; If it is determined to be at the edge of the load window, the real-time environmental disturbance index is integrated and matched with the historical working condition anti-disturbance model, the original working condition classification result is corrected and the working condition category priority is reordered, wherein the real-time environmental disturbance index includes sudden change in wind speed, sudden drop in temperature, and electromagnetic interference; The operating condition category corresponding to the highest confidence level in the comprehensive matching results is determined as the current operating condition category.

5. The electric energy comprehensive management and control method for high and low load conditions according to claim 1 is characterized in that: The calculating of the corresponding total system power demand value according to the operating condition category and the target power demand includes: Combined with the current operating condition category, a power demand record sequence matching the operating condition is extracted from historical data. Based on the power value changes in each period in the record, the average power level, maximum power surge amplitude, and power fluctuation duration at different stages are statistically analyzed to form a basic load characteristic description. Based on the task type, execution phase, and running time under the current working conditions, the target power demand is subdivided into multiple continuous time periods, and a corresponding load level is assigned to each time period. Then, the load level is matched with the corresponding power parameters in the basic load characteristic description to determine the preliminary target power demand sequence; Real-time collection of influencing factors related to the current operating environment, including wind speed, temperature, humidity, terrain undulation, and obstacle density. Based on the impact of different environmental factors on power consumption, the power value of each period in the preliminary target power demand sequence is adjusted step by step to obtain a target power sequence that includes environmental correction factors. Compare the current battery state of charge with the power buffer to assess whether the battery has the ability to handle a sudden load surge or release part of the base load. If the conditions are met, a corresponding load transfer correction value is introduced into the target power sequence to match the target power with the battery's response capability. The adjusted target power values ​​for each time period are combined with the basic power demand to form a set of final system total power demand values ​​in chronological order, which are used for the subsequent execution of power allocation and energy control strategies, so that the calculation results can fully reflect the comprehensive effects of working condition history, mission characteristics, environmental impact and battery capacity.

6. The electric energy comprehensive management and control method for high and low load conditions according to claim 1 is characterized in that: The determining of the output power of the power generation unit and the charge and discharge power of the battery according to the battery state of charge (SOC) value and the total power demand value of the system includes: According to the historical basic load model and current operating mode information corresponding to the operating condition label, the target power demand value of the system at the current moment is extracted, and the basic power part of the demand value is preliminarily set as the power output target of the power generation unit; Read the current SOC value of the battery and compare it with the preset SOC threshold range to determine whether it is in the charging zone, discharging zone or balance maintenance zone, and accordingly form an acceptable battery charge and discharge power adjustment range; If the current operating condition is normal, then on the premise of meeting the basic load demand, the remaining power is allocated to the battery for corresponding charging or discharging operations based on the interval position of the SOC value and the load change trend, and the combined allocation result of the power generation unit output power and the battery charging and discharging power is generated; If the current operating condition is an emergency condition and the power demand shows a sudden increase, the battery discharge power is set to the current maximum allowable value, subject to the constraint that it does not exceed the maximum rate discharge capability of the battery. The power of the power generation unit is temporarily maintained unchanged or slightly increased to stabilize the bus voltage. If a sudden drop in load is detected, the output power of the power generation unit will be quickly reduced, and the battery charging current limit model will be called up to set the battery charging power value without exceeding the maximum safe charging current of the battery. Finally, based on various judgment conditions, the target output power value of the power generation unit and the battery charging and discharging power value at the current moment will be output.

7. The electric energy comprehensive management and control method for high and low load conditions according to claim 1 is characterized in that: The controlling of the actual energy supply behavior of the power generation unit and the battery according to the output power of the power generation unit and the charge and discharge power of the battery includes: Receive the determined output power of the power generation unit and the battery charge and discharge power, and dynamically compare the sum of the two with the target bus power demand value to identify whether the current total energy supply meets the system's instantaneous power balance requirements, and generate an energy balance adjustment signal when insufficient or redundant energy is detected; Based on the energy balance adjustment signal, the dynamic adjustment limit of the power generation unit under the current mechanical speed, fuel supply and ambient temperature conditions is preferentially calculated to form a set of power generation adjustment boundary intervals, which are then output for the next power adjustment judgment; Based on the battery's current SOC value, temperature, aging level, and current cycle rate, the preset maximum charge current and maximum discharge rate parameters are queried to generate a battery regulation capability description vector. This vector is then compared with the aforementioned power generation adjustment boundary interval to derive a feasible multi-strategy energy supply combination for this round of regulation. Based on the current load type, predicted power fluctuation trends, and energy utilization priority rules, the optimal strategy is selected from all feasible energy supply combinations. If short-term load fluctuations are predicted to be frequent, battery response is prioritized; if the load is stable over a long period, power generation units are prioritized to provide basic energy. This determines the target execution output values ​​of the power generation units and batteries for each control cycle. The target execution output values ​​of the power generation unit and the battery are converted into corresponding control instruction parameters, including speed setting, excitation regulation, PWM control, and current loop control, and are respectively sent to the power generation unit and battery management system through the execution control module to drive them to perform actual energy supply operations, thereby realizing dynamic energy collaborative supply under closed-loop control based on power demand.

Images