A smart city base station power supply emergency supervision internet of things large model system and method

Abstract

The application provides a smart city base station power supply emergency supervision Internet of Things large model system and method, comprising an emergency supervision management platform, which is configured to: in response to detecting that power failure occurs in a region, determine at least one power failure region; for each of the at least one power failure region: according to base station data of the power failure region, determine backup power of the power failure region; according to power consumption characteristics of an emergency event in the power failure region, determine a predicted power consumption index of the emergency event; according to the backup power, basic power consumption and the predicted power consumption index of the power failure region, determine an estimated endurance time of the power failure region; according to the estimated endurance time and an estimated power supply time, generate a first scheduling instruction comprising a power supply time; based on the first scheduling instruction, control an automatic driving power supply vehicle to move to a base station associated with the power failure region at the power supply time, and control a generator of the automatic driving power supply vehicle to temporarily supply power to the base station. Through the system and method, efficient power supply scheduling can be achieved.

Description

Technical Field

[0001] This invention relates to the field of power supply technology, and in particular to a large-scale IoT model system and method for emergency monitoring of power supply for smart city base stations. Background Technology

[0002] Communication networks are a vital foundation for the normal operation of modern cities. In the event of a sudden emergency, ensuring emergency power supply to base stations is crucial. Existing emergency communication support technologies rely heavily on static data and human experience for passive responses, resulting in inaccurate prediction of backup power supply duration, delayed scheduling decisions, and inefficient resource allocation, making it difficult to cope with complex and ever-changing emergencies.

[0003] Therefore, we hope to provide a large-scale IoT model system and method for emergency monitoring of base station power supplies in smart cities, so as to achieve intelligent decision-making that combines accurate assessment with efficient scheduling. Summary of the Invention

[0004] This invention provides a smart city base station power emergency monitoring IoT big data model system, including an emergency monitoring management platform. The emergency monitoring management platform is configured to: in response to detecting a power outage in a region, identify at least one power outage area; for each of the at least one power outage area: determine the backup power supply for the power outage area based on base station data of the power outage area; determine a predicted power consumption sequence based on the power consumption characteristics and historical power consumption sequence of emergency events within the power outage area, using a power consumption prediction model, wherein the power consumption prediction model is a machine learning model, and the predicted power consumption sequence includes predicted power consumption of the emergency events at multiple moments within a future period; determine the estimated runtime of the power outage area based on the backup power supply, the base power consumption of the power outage area, and the predicted power consumption sequence; generate a first scheduling instruction including the power supply time based on the estimated runtime and estimated power supply time, and send it to the emergency monitoring perception and control platform; and control an autonomous driving power supply vehicle to move to a base station associated with the power outage area at the power supply time based on the first scheduling instruction, and control the generator of the autonomous driving power supply vehicle to temporarily power the base station.

[0005] This invention provides a method for emergency monitoring of base station power supply in smart cities. The method is executed by the emergency monitoring management platform of a smart city base station power supply emergency monitoring system, and includes: in response to detecting a power outage in a region, identifying at least one power outage region; for each of the at least one power outage region: determining the reserve power capacity of the power outage region based on base station data of the power outage region; determining a predicted power consumption sequence based on the power consumption characteristics and historical power consumption sequence of emergency events within the power outage region, using a power consumption prediction model, wherein the power consumption prediction model is a machine learning model, and the predicted power consumption sequence includes predicted power consumption of the emergency events at multiple moments within a future period; determining the estimated operating time of the power outage region based on the reserve power capacity, the base power consumption of the power outage region, and the predicted power consumption sequence; generating a first scheduling instruction including the power supply time based on the estimated operating time and the estimated power supply time, and sending it to the emergency monitoring perception and control platform; and controlling an autonomous driving power supply vehicle to move to a base station associated with the power outage region at the power supply time based on the first scheduling instruction, and controlling the generator of the autonomous driving power supply vehicle to temporarily supply power to the base station.

[0006] In some embodiments of this specification, by using machine learning models to predict power consumption, the estimated driving range is determined, and then the autonomous driving power supply vehicle is scheduled according to the estimated driving range and estimated power supply time to provide targeted temporary power supply to each base station, thus realizing intelligent decision-making that combines accurate assessment and efficient scheduling. Attached Figure Description

[0007] This specification will be further described by way of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting; in these embodiments, the same reference numerals denote the same structures, wherein:

[0008] Figure 1 This is a schematic diagram of the platform structure of a smart city base station power emergency monitoring IoT large-scale model system according to some embodiments of this specification; Figure 2 This is an exemplary flowchart of a smart city base station power emergency monitoring method according to some embodiments of this specification; Figure 3 This is an exemplary schematic diagram of a power consumption prediction model according to some embodiments of this specification; Figure 4 These are exemplary schematic diagrams of power supply prediction models shown in some embodiments of this specification; Figure 5 This is another exemplary flowchart of a smart city base station power emergency monitoring method according to some embodiments of this specification. Detailed Implementation

[0009] To more clearly illustrate the technical solutions of the embodiments in this specification, the accompanying drawings used in the description of the embodiments will be briefly introduced below. For those skilled in the art, these drawings can be applied to other similar scenarios without creative effort. The same reference numerals in the drawings represent the same structures or operations.

[0010] Flowcharts are used in this specification to illustrate the operations performed by the system according to embodiments of this specification. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, the steps can be processed in reverse order or simultaneously. Furthermore, other operations can be added to these processes, or one or more steps can be removed from them.

[0011] This specification provides a large-scale IoT model system and method for emergency power supply monitoring of smart city base stations, which can quickly respond to emergency events, dispatch autonomous driving power supply vehicles to perform emergency power supply to base stations in power outage areas, and ensure communication continuity.

[0012] Figure 1 This is a schematic diagram of the platform structure of a large-scale IoT model system for emergency monitoring of smart city base station power supply, based on some embodiments of this specification.

[0013] The Smart City Base Station Power Emergency Monitoring IoT Large-Scale Model System refers to a system composed of IoT model architectures that enables the efficient operation of large amounts of data. This system can integrate various artificial intelligence models to assist in data perception and processing.

[0014] In some embodiments, such as Figure 1 As shown, the smart city base station power emergency monitoring IoT large model system 100 may include an emergency monitoring user platform 110, an emergency monitoring service platform 120, an emergency monitoring management platform 130, an emergency monitoring sensor network platform 140, and an emergency monitoring perception and control platform 150.

[0015] The emergency monitoring user platform 110 refers to a platform that enables interaction with emergency monitoring users. In some embodiments, the emergency monitoring user platform can be configured as a terminal device.

[0016] The emergency monitoring service platform 120 refers to a platform that enables data communication between the emergency monitoring user platform 110 and the emergency monitoring management platform 130. In some embodiments, the emergency monitoring service platform can be configured as a communication device or a gateway.

[0017] The emergency monitoring and management platform 130 refers to a platform used to provide computing, networking, and storage capabilities. In some embodiments, the emergency monitoring and management platform can be configured as a server. In some embodiments, the server can be regional or remote. In some embodiments, the server can be implemented on a cloud platform or provided virtually.

[0018] In some embodiments, the emergency monitoring and management platform 130 is configured to: in response to detecting a power outage in a region, identify at least one power outage region; for each of the at least one power outage region: determine the backup power supply of the power outage region based on base station data of the power outage region; determine the predicted power consumption index of the emergency event based on the power consumption characteristics of the emergency event in the power outage region; determine the estimated endurance time of the power outage region based on the backup power supply, the base power consumption of the power outage region, and the predicted power consumption index; generate a first scheduling instruction including the power supply time based on the estimated endurance time and the estimated power supply time, and send it to the emergency monitoring and perception control platform; and control an autonomous driving power supply vehicle to move to the base station associated with the power outage region at the power supply time based on the first scheduling instruction, and control the generator of the autonomous driving power supply vehicle to temporarily supply power to the base station.

[0019] The emergency monitoring sensor network platform 140 refers to a platform that enables data communication between the emergency monitoring management platform 130 and the emergency monitoring perception and control platform 150. In some embodiments, the emergency monitoring sensor network platform can be configured as a communication device or a gateway.

[0020] The emergency monitoring and control platform 150 refers to an execution platform used to perform emergency monitoring and power supply. In some embodiments, the emergency monitoring and control platform may include hardware devices for performing emergency monitoring and power supply, such as autonomous driving power supply vehicles, IoT sensors, base stations, and backup power supplies.

[0021] An autonomous power supply vehicle is a vehicle equipped with an emergency power supply and autonomous driving capabilities, which can move to a designated location to provide temporary power to base stations.

[0022] IoT sensors are sensors used to monitor power-related data within a base station and transmit it to the Internet of Things (IoT). For example, IoT sensors may include voltage sensors, current sensors, or oil level sensors.

[0023] In some embodiments, the smart city base station power emergency monitoring IoT big data model system 100 also includes an IoT model architecture. The IoT model architecture is used to enable the efficient operation of large amounts of data within the system. In some embodiments, AI models can be applied to the IoT model architecture to assist in data perception and processing.

[0024] In some embodiments, the emergency monitoring and management platform 130 includes a data center. The IoT model architecture can be located within the data center. The data center also includes a database and a computing unit. The database stores data obtained by the emergency monitoring and management platform 130 from interactions with other platforms and / or processing results or instructions generated by the emergency monitoring and management platform 130. The computing unit performs data processing, and can call the corresponding model from the IoT model architecture and retrieve the corresponding data to be processed from the database.

[0025] It should be noted that the above description of the smart city base station power emergency monitoring IoT large-scale model system 100 and its platform should not limit this specification to the scope of the embodiments described. Those skilled in the art, after understanding the principles of the system, may arbitrarily combine the various platforms or construct subsystems connected to other platforms. For example, the platforms may share a single storage module, or each platform may have its own independent storage module. Such modifications are all within the scope of this specification.

[0026] This specification provides a smart city base station power emergency monitoring method, comprising: in response to detecting a power outage in a region, identifying at least one power outage area; for each of the at least one power outage area: determining the backup power capacity of the power outage area based on base station data of the power outage area; determining a predicted power consumption sequence based on the power consumption characteristics and historical power consumption sequence of emergency events in the power outage area, using a power consumption prediction model, wherein the power consumption prediction model is a machine learning model and the predicted power consumption sequence includes the predicted power consumption of emergency events at multiple moments in the future period; determining the estimated endurance time of the power outage area based on the backup power capacity, the basic power consumption of the power outage area, and the predicted power consumption sequence; generating a first scheduling instruction including the power supply time based on the estimated endurance time and the estimated power supply time, and sending it to an emergency monitoring perception and control platform; and controlling an autonomous driving power supply vehicle to move to a base station associated with the power outage area at the power supply time based on the first scheduling instruction, and controlling the generator of the autonomous driving power supply vehicle to temporarily supply power to the base station.

[0027] Figure 2 This is an exemplary flowchart illustrating a smart city base station power emergency monitoring method according to some embodiments of this specification. Figure 2 As shown, process 200 includes the following steps. In some embodiments, process 200 may be executed by the emergency monitoring management platform 130 of the smart city base station power emergency monitoring IoT big data model system 100.

[0028] Step 210: In response to the detection of a power outage in a region, at least one power outage region is identified.

[0029] A power outage area refers to the geographical or administrative region where a power outage is detected. For example, a power outage area could be a street, a community, or the coverage area of ​​a specific base station. Alternatively, a power outage area could be a specific area where a power outage has occurred, defined by the power supply facilities.

[0030] Regions can be defined by geographical location or administrative division, or by corresponding base stations. For example, the geographical area or administrative region covered by a single base station can be defined as a region.

[0031] In some embodiments, the emergency monitoring and management platform can obtain information from external data regarding whether power outages have occurred in various regions. If a power outage is detected, the platform queries the external data to identify the affected region. External data refers to real-time data obtained from departments such as emergency management, power, and meteorology, and may include the nature of the power outage event (e.g., typhoon, earthquake, large-scale power outage), its scale, and the estimated time for mains power restoration.

[0032] In other embodiments, the emergency monitoring and management platform can identify areas where the main power supply voltage / current is 0 and the backup power supply voltage / current is not 0 as power outage areas.

[0033] In some embodiments, the emergency monitoring and management platform can also determine the power outage area through various other methods. For example, the platform can determine the power outage area based on power outage information provided by users through the emergency monitoring user platform.

[0034] Step 220: Perform steps 221 to 225 for each of at least one power outage area.

[0035] Step 221: Determine the backup power supply for the power outage area based on the base station data of the power outage area.

[0036] Base station data refers to the operational data of the backup power supply of base stations associated with the power outage area. Backup power supplies can include battery packs, generators, etc. For example, base station data may include voltage, current, battery internal resistance, and generator fuel level. Base station data can be collected using IoT sensors.

[0037] Backup power refers to the total amount of electricity that the backup power supply of the base station associated with the power outage area is expected to provide after a power outage.

[0038] In some embodiments, the emergency monitoring and management platform can determine the backup power supply for the power outage area through various methods based on base station data in the outage area.

[0039] For example, the emergency monitoring and management platform can determine multiple clustering vectors based on historical base station data from different regions. The label of each clustering vector represents the actual reserve power of the corresponding region under the historical conditions corresponding to that clustering vector. Then, based on the base station data of the current power outage area, a target vector is determined. The multiple clustering vectors and the target vector are clustered to obtain multiple clusters. The cluster containing the target vector is denoted as the target cluster. The average of the labels corresponding to all clustering vectors in the target cluster is used as the reserve power corresponding to the target vector.

[0040] For example, the emergency monitoring and management platform can construct a first preset table based on the base station data and reserve power of a single base station in historical records, and query the first preset table based on the base station data to determine the reserve power corresponding to each base station. The reserve power corresponding to the base stations associated with the power outage area is summed up as the reserve power of the power outage area.

[0041] The first preset table contains the correspondence between base station data and reserve power for a single base station, determined by historical records and experience.

[0042] In some embodiments, the emergency monitoring and management platform can also determine the reserve power through various other methods. For example, the platform can also determine the reserve power based on data such as voltage, current, and battery internal resistance monitored by IoT sensors, using algorithms such as the Coulomb integration method and the internal resistance method.

[0043] Step 222: Determine the predicted power consumption index of the emergency event based on the power consumption characteristics of the emergency event within the power outage area.

[0044] An emergency event is an event that occurs within the area affected by a power outage and requires additional power supply. Examples of emergency events include concerts, sporting events, or other large-scale temporary events.

[0045] Power consumption characteristics refer to data used to describe the power consumption characteristics of emergency events. For example, power consumption characteristics may include the number, type, and scale of emergency events.

[0046] Predicted power consumption metrics refer to the power consumption metrics related to emergency events estimated over a future period, such as the total expected power consumption and the expected duration of power consumption.

[0047] A future time period refers to a period of time extending forward from the current moment. For example, a future time period could be an hour, a day, etc.

[0048] In some embodiments, the predicted power consumption metric may include the average power consumption of an emergency event over a future period.

[0049] In some embodiments, the emergency monitoring and management platform can determine the predicted power consumption index of an emergency event through various methods based on the power consumption characteristics of the emergency event within the power outage area.

[0050] For example, the emergency monitoring and management platform can match the first feature vector in the first vector database, and take the first reference vector with the highest similarity as the first target vector; and determine the label corresponding to the first target vector as the predicted power consumption index of the emergency event.

[0051] The first feature vector represents the power consumption characteristics of the emergency event to be matched. The first vector database includes a first reference vector and its corresponding label. The first reference vector is composed of the historical power consumption characteristics of multiple historical emergency events. The corresponding label can be determined based on the actual average power consumption of the historical emergency event during the subsequent preset time period (such as 3 hours, 6 hours, etc.) corresponding to the first reference vector. Similarity can be calculated based on cosine distance and Euclidean distance.

[0052] For example, the emergency monitoring and management platform can construct a second preset table based on the power consumption characteristics and actual power consumption of emergency events in historical records, and query the second preset table based on the power consumption characteristics of the current emergency event to determine the corresponding predicted power consumption index.

[0053] The second preset table contains the correspondence between the power consumption characteristics of emergency events and the predicted power consumption indicators, which is determined by historical records and experience.

[0054] In some embodiments, the emergency monitoring and management platform may also determine the predicted power consumption indicators through other means. For example, the predicted power consumption indicators may include a predicted power consumption sequence, which the emergency monitoring and management platform can determine using a power consumption prediction model, see [link to relevant documentation]. Figure 3 And the following explanation.

[0055] Step 223: Determine the estimated battery life of the power outage area based on the backup power, the basic power consumption of the outage area, and the predicted power consumption indicators.

[0056] Baseline power consumption refers to the normal power consumption of the outage area when no emergency event occurs. In some embodiments, the baseline power consumption of the outage area can be obtained by statistically analyzing the average power consumption during periods without emergency events in historical data.

[0057] Estimated battery life refers to the duration for which a backup power supply is expected to provide continuous power under current load conditions.

[0058] In some embodiments, the emergency monitoring and management platform can determine the estimated battery life of a power outage area through various methods based on the backup power supply, the basic power consumption of the outage area, and the predicted power consumption indicators.

[0059] For example, the emergency monitoring and management platform can determine the estimated battery life based on the basic power consumption of the power outage area per unit time and the predicted power consumption index of the emergency event, combined with the backup power. For instance, the estimated battery life can be calculated based on formula (1): (1) This indicates the estimated battery life. Indicates reserve power. This represents the base power consumption of the area affected by the power outage per unit time. This represents the predicted power consumption per unit of time.

[0060] For example, the emergency monitoring and management platform can match the second feature vector in the second vector database, and take the second reference vector with the highest similarity as the second target vector; and take the label corresponding to the second target vector as the estimated battery life.

[0061] The second feature vector consists of the reserve power to be matched, the base power consumption of the power outage area, and the predicted power consumption index. The second vector database includes a second reference vector and its corresponding label. The second reference vector consists of the historical actual reserve power, the base power consumption of the power outage area, and the predicted power consumption index. Its corresponding label is determined based on the subsequent actual battery life corresponding to the second reference vector. Similarity can be calculated based on cosine distance and Euclidean distance.

[0062] In some embodiments, the emergency monitoring and management platform may also determine the estimated battery life through other means. For example, the emergency monitoring and management platform may determine the estimated battery life through a power consumption prediction model, see [link to relevant documentation]. Figure 3 And the following explanation.

[0063] Step 224: Based on the estimated battery life and estimated power supply time, generate the first dispatch instruction including the power supply time and send it to the emergency monitoring and sensing control platform.

[0064] Estimated power supply time refers to the time from the current moment until the expected restoration of mains power to the affected area.

[0065] In some embodiments, the estimated power supply time can be obtained from external data. For example, if the power company estimates that the power outage will be repaired in 2 hours, then the estimated power supply time is 2 hours.

[0066] The first dispatch instruction refers to the instruction used to direct and schedule the autonomous power supply vehicle to provide power. For example, the first dispatch instruction may include the base station coordinates, the power supply time, and the identification information of the autonomous power supply vehicle.

[0067] The power supply time refers to the expected start time of temporary power supply. In some embodiments, the power supply time can be the start time of a power supply period, which is the period between the current time plus the estimated battery life and the current time plus the estimated power supply time.

[0068] Understandably, it is only necessary to dispatch an autonomous driving power vehicle for temporary power supply when the estimated driving range is less than the estimated power supply time.

[0069] In some embodiments, the emergency monitoring and management platform can generate a first dispatch instruction, including the power supply time, in various ways based on the estimated battery life and estimated power supply time.

[0070] For example, the emergency monitoring and management platform can select N base stations closest to the power outage area as base stations associated with the power outage area, and M autonomous power supply vehicles as autonomous power supply vehicles to perform power supply. The coordinates of these base stations, the identification information of the autonomous power supply vehicles, and the power supply time are integrated into the command as the first dispatch command.

[0071] For example, the emergency monitoring and management platform can use a path planning algorithm to determine the base stations and autonomous power supply vehicles that can provide power the fastest, based on the locations of N selected base stations and M autonomous power supply vehicles. It can then integrate the coordinates of these base stations, the identification information of the autonomous power supply vehicles, and the power supply time into the command to obtain the first dispatch command.

[0072] Step 225: Based on the first scheduling instruction, control the autonomous driving power supply vehicle to move to the base station associated with the power outage area at the power supply time, and control the generator of the autonomous driving power supply vehicle to provide temporary power to the base station.

[0073] Base stations associated with the power outage area refer to base stations that can provide radio coverage to the power outage area.

[0074] In some embodiments, the emergency monitoring and management platform can, based on the identification information of the autonomous power supply vehicle in the first dispatch instruction, control the corresponding autonomous power supply vehicle to move to the base station associated with the power outage area in the first dispatch instruction at the power supply time, and control the generator of the autonomous power supply vehicle to provide temporary power to the base station at a standard speed. The standard speed can be preset.

[0075] In some embodiments of this specification, the power outage area is determined by detection, and a first scheduling instruction is generated to control the autonomous driving power supply vehicle to provide temporary power supply. This integrates power outage detection, backup power determination, power consumption prediction, and emergency power supply vehicle scheduling, thereby realizing automated early warning and emergency response to base station power shortages.

[0076] Figure 3 This is an exemplary schematic diagram of a power consumption prediction model shown according to some embodiments of this specification.

[0077] In some embodiments, such as Figure 3 As shown, the emergency monitoring and management platform is further configured to: determine the predicted power consumption sequence 340 based on the power consumption characteristics 310 and historical power consumption sequence 320 of the emergency event through the power consumption prediction model 330; and determine the estimated battery life 370 based on the backup power 350, the basic power consumption 360 and the predicted power consumption sequence 340.

[0078] Historical power consumption sequence 320 refers to a sequence of actual power consumption data of base stations associated with the power outage area at multiple moments within a historical period. The historical period can refer to a period of time prior to the current moment. For example, if the historical period is the past 24 hours and the time interval between multiple moments is 15 minutes, the historical power consumption sequence 320 can include the actual power consumption sequence of the base station every 15 minutes within the past 24 hours.

[0079] The predicted power consumption sequence 340 refers to the relevant power consumption metrics at multiple moments within an estimated future time period. In some embodiments, the predicted power consumption sequence 340 includes predicted power consumption for emergency events at multiple moments within the future time period. For example, the predicted power consumption sequence 340 may be a sequence including expected power consumption values ​​every 15 minutes within the future time period. In some embodiments, averaging the predicted power consumption sequence 340 can determine an average power consumption metric.

[0080] The power prediction model 330 refers to a model used to determine a predicted power consumption sequence. In some embodiments, the power prediction model 330 may include a neural network model and other trained machine learning models, such as a Long Short-Term Memory (LSTM) model.

[0081] In some embodiments, the power consumption prediction model can be trained based on a large number of first training samples with a first label. The first training samples can be the power consumption characteristics and power consumption sequences of sample emergency events in historical records. The power consumption sequence is the power consumption sequence of sample base stations associated with the power outage area where the sample emergency event is located during historical periods. The first label can be the historical actual power consumption sequence of multiple moments in the future period corresponding to the sample base station. The first label can be automatically labeled by the emergency monitoring and management platform based on historical records.

[0082] In some embodiments, the emergency monitoring and management platform can perform multiple rounds of iterative training to obtain a power consumption prediction model. One iteration includes: inputting one or more first training samples into an initial power consumption prediction model to obtain outputs corresponding to the one or more first training samples; substituting the outputs of the initial power consumption prediction model and the corresponding first labels into a predefined first loss function; calculating the value of the first loss function; and iteratively updating the initial power consumption prediction model based on the first loss function. For example, updating can be based on gradient descent. When the value of the loss function meets the iteration completion condition, training is complete, and a trained power consumption prediction model is obtained. The iteration completion condition may include the convergence of the first loss function, the number of iterations reaching a threshold, etc.

[0083] In some embodiments, such as Figure 3 As shown, the input to the power consumption prediction model 330 also includes historical state features 380.

[0084] Historical status feature 380 refers to data related to the operational status of base stations associated with the power outage area where the emergency event occurred during a historical period. For example, historical status feature 380 may include historical power efficiency 381, historical ambient temperature 382, ​​and historical equipment load rate 383.

[0085] Historical power efficiency (381) refers to the power supply efficiency of a power device over a historical period. For example, historical power efficiency can be the average energy conversion ratio of a battery during its charge-discharge cycles over a historical period.

[0086] In some embodiments, the emergency monitoring and management platform can obtain the total power consumption of the base station and the actual power consumption of the downstream equipment through the power meters of the base station and the power meters of the downstream equipment, and determine the historical power efficiency by the ratio of the actual power consumption of the downstream equipment to the total power consumption of the base station.

[0087] Historical ambient temperature 382 refers to the ambient temperature at which backup power supplies or other equipment were located during a historical period. In some embodiments, the emergency monitoring and management platform can obtain historical ambient temperatures through temperature sensors installed in the base station equipment room or equipment enclosure.

[0088] Historical equipment load factor 383 refers to the ratio of the actual operating power of equipment to its rated maximum power during a historical period. For example, if a device with a rated power of 10kW operated at 5kW in the past hour, its historical equipment load factor is 50%.

[0089] In some embodiments, when the input of the power consumption prediction model includes historical state features, the first training sample may further include sample state features, which are the operating state features of the sample base station in the period preceding the time corresponding to the first training sample.

[0090] In some embodiments of this specification, by also using historical state characteristics as input to the power consumption prediction model, the power consumption prediction takes into account the influence of the environment and device state, thereby further improving the accuracy of the prediction.

[0091] In some embodiments, the emergency monitoring and management platform can determine the estimated battery life 370 in various ways based on the backup power 350, the basic power consumption 360, and the predicted power consumption sequence 340.

[0092] For example, the emergency monitoring and management platform can accumulate the basic power consumption and the predicted power consumption sequence until the total power consumption after accumulation is greater than or equal to the backup power, and use the time at this point as the estimated battery life.

[0093] For example, assuming the predicted power consumption sequence is (A1, A2, A3, ..., An), the base power consumption per unit time is B, and the time interval is t, the emergency monitoring and management platform can determine the estimated battery life through the following steps: 1) Initialize i=1, where i represents the iteration number, the estimated battery life is 0, and the power consumption is 0. Execute the following iterations: 2) Update power consumption. Updated power consumption = power consumption + Ai t+B t; 3) Determine if the updated power consumption is greater than or equal to the reserve power: 4.1) If not, update the estimated battery life. The updated estimated battery life = estimated battery life + t, i = i + 1, and return to step 2) to enter the next iteration. 4.2) If so, end the iteration and take the estimated battery life at this time as the final estimated battery life.

[0094] For example, the emergency monitoring and management platform can match the third feature vector in the third vector database, and take the third reference vector with the highest similarity as the third target vector; and take the label corresponding to the third target vector as the estimated battery life.

[0095] The third feature vector consists of the sequence of reserve power, base power consumption, and predicted power consumption to be matched. The third vector database includes a third reference vector and its corresponding label. The third reference vector consists of the historical actual reserve power, base power consumption, and predicted power consumption sequence, and its corresponding label can be determined based on the subsequent actual battery life corresponding to the third reference vector. Similarity can be calculated based on cosine distance and Euclidean distance.

[0096] In some embodiments of this specification, by introducing a power consumption prediction model to determine the predicted power consumption sequence, the prediction of a single predicted power consumption indicator (e.g., average predicted power consumption) is upgraded to the prediction of a more refined predicted power consumption sequence, thereby more accurately determining the estimated battery life of the backup power supply.

[0097] In some embodiments, the time intervals between multiple moments in the predicted power consumption sequence are related to the length of the confidence interval for the estimated power supply time.

[0098] The confidence interval for estimated power supply time represents the time range within which the estimated power supply time may fall. The length of the confidence interval for estimated power supply time is the difference between the upper and lower boundaries of the confidence interval. For example, if the confidence interval is [6 hours, 8 hours], the length of the confidence interval is 2 hours.

[0099] In some embodiments, the emergency monitoring and management platform can obtain multiple possible estimated mains power restoration times from external data, and determine the length of the confidence interval for the estimated power supply time as the difference between the minimum and maximum values ​​among these multiple possible estimated mains power restoration times. If the multiple possible estimated power supply times obtained from external data are all the same, the length of the confidence interval for the estimated power supply time can be determined as a preset length.

[0100] In some embodiments, the time intervals between multiple moments in the predicted power consumption sequence may be positively correlated with the length of the confidence interval for the estimated power supply time.

[0101] When the confidence interval for estimated power supply time is narrow, the prediction result is relatively certain. In this case, the scheduling window is narrow, and the margin for error in decision-making is small. A 15-minute overestimation or underestimation of the power supply duration can determine the success or failure of scheduling. Therefore, high-precision, fine-grained information is urgently needed for decision-making. Conversely, when the confidence interval for estimated power supply time is wide, the biggest risk source is the uncertainty of the power restoration time itself. Therefore, using a larger time interval can smooth out short-term noise, more clearly show long-term power consumption patterns, and significantly reduce computational resource consumption. By positively correlating the time intervals of multiple moments with the length of the confidence interval for estimated power supply time, actual needs and the computational resources of the emergency monitoring and management platform can be comprehensively considered to ensure the smooth operation of emergency power supply.

[0102] In some embodiments, the emergency monitoring and management platform is further configured to: generate monitoring and control instructions according to time intervals and send them to the emergency monitoring and sensing control platform; control the Internet of Things sensors in the power outage area to collect data at a sampling rate based on the monitoring and control instructions, and upload the collected data at an upload frequency.

[0103] Monitoring and control commands are instructions used to adjust the monitoring parameters of monitoring equipment. For example, monitoring and control commands may include instructions requiring IoT sensors to increase data sampling rate and upload frequency.

[0104] The sampling rate refers to the amount of data collected by an IoT sensor each time it performs data acquisition. For example, a sampling rate of n can mean that the IoT sensor collects n data points each time it acquires data.

[0105] Upload frequency refers to the rate at which the sensor transmits the collected data to the management platform. For example, the upload frequency could be once every minute.

[0106] In some embodiments, the emergency monitoring and management platform can generate monitoring and control instructions in various ways based on time intervals.

[0107] For example, the emergency monitoring and management platform can construct a third preset table based on the time intervals, sampling volume, and upload frequency in historical records, and query the third preset table based on the currently determined time interval to determine the corresponding sampling volume and upload frequency, thereby determining the monitoring and control instructions.

[0108] The third preset table contains the correspondence between the range of time intervals, the number of samples, and the upload frequency. This correspondence can be determined based on historical records and experience. For example, when the time interval is short, the number of samples can be large and the upload frequency can be high; when the time interval is long, the number of samples can be small and the upload frequency can be low.

[0109] For example, the emergency monitoring and management platform can determine the sampling volume and upload frequency of IoT sensors based on a preset formula according to the time interval, thereby generating corresponding monitoring and control instructions. For instance, the preset formula can be shown in equation (2) below: (2) Indicates time interval, This indicates the duration between each data collection session corresponding to the upload frequency. For preset coefficients, Indicates the sampling quantity. This indicates the baseline sampling quantity. The preset coefficient and the baseline sampling quantity can be preset by technical personnel.

[0110] In some embodiments, the emergency monitoring and management platform can also generate monitoring and control instructions through various other means.

[0111] In some embodiments, the emergency monitoring and management platform can control the Internet of Things (IoT) sensors in the power outage area to collect data at a sampling rate and upload the collected data to the emergency monitoring and management platform at an upload rate, based on the sampling rate and upload frequency in the monitoring and control instructions.

[0112] In some embodiments of this specification, the sampling amount and upload frequency of the IoT sensor are adjusted according to the time interval. Appropriate adjustments can be made according to actual needs to achieve dynamic and intelligent control of monitoring resources.

[0113] Figure 4 This is an exemplary schematic diagram of a power supply prediction model according to some embodiments of this specification.

[0114] In some embodiments, such as Figure 4 As shown, the emergency monitoring and management platform is further configured to: construct a power supply map 430 based on base station data 410 and location relationships 420 from multiple regions; determine the confidence interval 450 of the estimated power supply time using a power supply prediction model 440 based on the power supply map 430; and determine the estimated power supply time 460 based on the confidence interval 450 of the estimated power supply time. (See also: [link to base station data 410]). Figure 2 And the relevant descriptions mentioned above.

[0115] Location relationship 420 refers to the spatial relationship between different entities or regions. For example, location relationship 420 can include the adjacency between regions, the distance between base stations, or whether they are in the same geographical region.

[0116] The power supply map 430 is a structured map containing multiple nodes and edges, used to describe the power supply or geographical connectivity between different base stations. For example, nodes in the power supply map can represent base stations, and node features can include the base station's operating status (e.g., whether it is working properly) and base station data; if two base stations are connected by a power supply line, or if the areas where the two base stations are located are adjacent (geographically bordering), then there is an edge between the two base station nodes in the power supply map, and the edge feature can be the distance between the two base stations.

[0117] The power supply prediction model 440 is a model used to determine a confidence interval for the estimated power supply time. In some embodiments, the power supply prediction model includes a machine learning model. For example, the power supply prediction model may include one or a combination of graph neural network (GNN) models, deep neural network (DNN) models, etc.

[0118] In some embodiments, such as Figure 4 As shown, the power supply prediction model 440 may include a power outage feature extraction layer 441 and a power outage time prediction layer 443. The power outage feature extraction layer 441 may be a GNN, and the power outage time prediction layer 443 may be a DNN.

[0119] In some embodiments, the input to the power supply prediction model 440 may also include electrical fault features 444.

[0120] In some embodiments, the input of the power outage feature extraction layer 441 includes a power supply map 430, and the output includes power supply map features 442. The input of the power outage time estimation layer 443 includes power supply map features 442 and electrical fault features 444, and the output includes a confidence interval 450 for the estimated power supply time.

[0121] Power supply map feature 442 may include the scale and type features of the power outage area.

[0122] Scale characteristics reflect the scale of the power outage in the affected area; for example, they may include the number of outage nodes in the subgraph and the total number of affected users.

[0123] Type characteristics include features reflecting the outage types of different outage areas. For example, these can include the density, average path length, and betweenness centrality of multiple subgraphs in the power supply map. Each subgraph corresponds to an outage area, and the average path length refers to the average length of the edges between outage nodes within the subgraph. A high subgraph density may indicate a failure of core equipment such as regional transformers, corresponding to a clustered outage. A node failure with a long average path length and high betweenness centrality indicates a problem with the main power line, corresponding to a chained outage.

[0124] Electrical fault feature 444 is data reflecting the electrical characteristics of a period of time prior to a power outage. In some embodiments, the emergency monitoring and management platform can preprocess base station data from a period of time prior to the power outage to obtain the electrical fault features. Preprocessing may refer to Fourier transform, wavelet analysis, etc.

[0125] In some embodiments, the power supply prediction model 440 can be trained using a second training sample and a second label. The second training sample may include sample power supply maps and sample electrical fault features corresponding to multiple historical time points. The second label may be the range of subsequent actual power supply times for multiple historical power outages corresponding to the second training sample.

[0126] In some embodiments, the emergency monitoring and management platform can jointly train the power outage feature extraction layer 441 and the power outage time estimation layer 443 to obtain a power supply prediction model 440. For example, the emergency monitoring and management platform inputs the sample power supply map into the power outage feature extraction layer of the initial power supply prediction model, inputs the output of the power outage feature extraction layer and the sample electrical fault features into the power outage time estimation layer of the initial power supply prediction model, and constructs a second loss function based on the output of the power outage time estimation layer and the second label. Then, the emergency monitoring and management platform iteratively updates the parameters of the power outage feature extraction layer and the power outage time estimation layer in the initial power supply prediction model based on the second loss function until a preset condition is met, thus obtaining a trained power supply prediction model. The preset condition may be the convergence of the second loss function or the training period reaching a threshold.

[0127] In some embodiments, the emergency monitoring and management platform can determine the estimated power supply time 460 in various ways based on the confidence interval 450 of the estimated power supply time.

[0128] For example, an emergency monitoring and management platform can determine the estimated power supply time as the median of the confidence interval for the estimated power supply time. As another example, the platform can determine the estimated power supply time as the maximum value of the confidence interval for the estimated power supply time.

[0129] In some embodiments of this specification, by constructing a power supply map and using a jointly trained power supply prediction model to process the power supply map to determine the confidence interval of the estimated power supply time, the system can autonomously determine the estimated power supply time even when external data is lacking, thereby enhancing the robustness and information self-sufficiency of the system.

[0130] Figure 5 This is another exemplary flowchart of a smart city base station power emergency method according to some embodiments of this specification.

[0131] like Figure 5 As shown, the emergency monitoring and management platform can execute steps 510 to 540.

[0132] Step 510: Determine the power supply priority for at least one power outage area based on the regional importance of the at least one power outage area, the level of event impact, and the urgency of power supply.

[0133] Regional importance is an indicator that measures the importance of social services in a specific geographical area. In some embodiments, the emergency monitoring and management platform can determine regional importance based on the number of important buildings (hospitals, schools, etc.) within the power outage area. For example, each type of important building can correspond to a preset importance value. The emergency monitoring and management platform can multiply the importance values ​​of different important buildings within the power outage area by the corresponding number of buildings, and add the products corresponding to different types of important buildings to obtain the regional importance; the higher the value, the more important the area.

[0134] The event impact level reflects the degree of social influence of an emergency event. In some embodiments, the emergency monitoring and management platform can pre-set an impact index for each type of emergency event, and determine the impact value of the emergency event based on the corresponding impact index and scale multiplier. The scale multiplier is an indicator reflecting the size of the emergency event; the larger the scale, the larger the scale multiplier.

[0135] In some embodiments, the emergency monitoring and management platform normalizes the impact values ​​of all emergency events within the power outage area (e.g., Min-Max normalization) and then performs a weighted summation to obtain the event impact level; the closer the level is to 1, the greater the social impact. The weights for the weighted summation can be set based on experience.

[0136] Battery life urgency is an indicator reflecting the degree of urgency required for power supply. In some embodiments, the emergency monitoring and management platform may take the reciprocal of the absolute value of the difference between the estimated battery life and the estimated power supply time as the battery life urgency value, with a higher value indicating greater urgency.

[0137] Power supply priority refers to the order in which rescue or power supply is prioritized for each power outage area.

[0138] In some embodiments, the emergency monitoring and management platform can determine the power supply priority of the power outage area through various methods based on the regional importance of the power outage area, the level of impact of the event, and the urgency of power resumption.

[0139] For example, the emergency monitoring and management platform can normalize the regional importance, event impact level, and urgency of power restoration for each power outage area, and then perform a weighted summation to obtain a score corresponding to each power outage area, which serves as the corresponding power supply priority. The higher the score, the higher the priority. The weighting coefficients for the weighted summation can be preset.

[0140] In some embodiments, the emergency monitoring and management platform can also determine the power supply priority of the power outage area through various other methods. For example, the emergency monitoring and management platform can construct a fourth preset table, which contains the correspondence between regional importance, event impact level, power urgency, and power supply priority. The power supply priority of the power outage area can be determined by querying the fourth preset table.

[0141] Step 520: Determine the target power supply strategy for at least one power outage area based on the power supply time, power supply priority, and location data of the autonomous driving power supply vehicle for at least one power outage area.

[0142] The location data of autonomous driving power supply vehicles reflects the locations of multiple autonomous driving power supply vehicles. For example, it may include identification information and corresponding location coordinates of multiple autonomous driving power supply vehicles. In some embodiments, the emergency monitoring and management platform can obtain the location data of autonomous driving power supply vehicles through an emergency monitoring perception and control platform (such as a positioning device).

[0143] A target power supply strategy refers to a comprehensive power supply plan developed to provide temporary power to an area experiencing a power outage. In some embodiments, the target power supply strategy may include an autonomous power vehicle (i.e., the target power supply vehicle) used for power supply, a target power supply path, and a target power supply rotation speed when supplying power to a base station. The target power supply rotation speed may be a constant value or multiple values ​​corresponding to multiple times.

[0144] In some embodiments, the emergency monitoring and management platform can determine the target power supply strategy for the power outage area based on the power supply time, power supply priority, and location data of the autonomous power supply vehicle in the power outage area, using path planning algorithms (such as simulated annealing algorithm, genetic algorithm, etc.).

[0145] For example, the emergency monitoring and management platform can select one or more autonomous power supply vehicles that are close to the power outage area as the initial power supply vehicle based on the location data of the power outage area and the autonomous power supply vehicle. Using a path planning algorithm, it can perform multiple iterations based on a cost function to determine the target power supply path.

[0146] For example, using a path planning algorithm to perform multiple iterations based on a cost function may include the following steps: 1) Generate candidate paths: Generate candidate paths based on the coordinates of the initial power supply vehicle and the base station. For example, candidate paths can be generated for each initial power supply vehicle based on the proximity principle. Alternatively, candidate paths for all initial power supply vehicles can be generated randomly.

[0147] 2) Determine the total travel cost: Calculate the sum of the travel times of all initially powered vehicles to the corresponding power outage areas, and use this as the total travel cost.

[0148] 3) Determine the total time cost: Calculate the time from the dispatch of the first vehicle to the arrival of the last vehicle at the corresponding base station, and use this as the total time cost.

[0149] 4) Determine the total priority penalty: Each base station corresponds to a priority penalty value. For example, the priority penalty corresponding to a base station can be calculated based on the following formula (3): (3) For base stations Priority penalty value, This represents the initial total number of vehicles powered. This indicates the arrival time of the initial power supply vehicle at the base station when using the candidate path. Duration Indicates base station The corresponding duration between the current moment and the power supply moment, Indicates the base penalty value. Indicates base station The corresponding priority weight. The base penalty value can be a preset constant, such as 1,000,000. The priority weight of the base station can be determined according to the power supply priority of the power outage area corresponding to the base station. The higher the power supply priority, the lower the corresponding priority weight can be. For example, the reciprocal of the power supply priority can be taken as the priority weight.

[0150] The total priority penalty is the sum of the priority penalty values ​​corresponding to all base stations.

[0151] 5) Calculate the cost function: By taking a weighted average of the total travel cost, total time cost and priority penalty, the value of the cost function for the corresponding candidate path is determined. The weights of the weighted average can be preset.

[0152] The emergency monitoring and management platform can repeat the above steps multiple times until the iteration termination condition is met, and then determine the candidate path with the minimum cost function calculation result as the target power supply path. For example, the iteration termination condition could be the convergence of the cost function or reaching the target number of iterations.

[0153] In some embodiments, the emergency monitoring and management platform determines the target power supply speed based on the predicted power consumption (predicted power consumption index or predicted power consumption sequence) of the power outage area using a preset algorithm.

[0154] For example, a preset algorithm divides the predicted power consumption index by the generator's torque to obtain the target power supply speed. Another preset algorithm divides the predicted power consumption of emergency events at multiple times in the predicted power consumption sequence by the generator's torque to obtain the corresponding power supply speeds at those times, which are then used as the target power supply speeds.

[0155] In some embodiments, the emergency monitoring and management platform can also determine the target power supply strategy through other means. For example, the emergency monitoring and management platform can utilize multi-objective optimization algorithms (such as particle swarm optimization, multi-objective Bayesian optimization, etc.) to determine the target power supply vehicle, the target power supply path, and the target power supply rotation speed.

[0156] Step 530: Generate a second dispatch instruction based on the target power supply strategy and send it to the emergency monitoring and sensing control platform.

[0157] The second scheduling instruction refers to the instruction generated according to the target power supply strategy to guide the autonomous driving power supply vehicle in providing power. For example, the second scheduling instruction may include the identification information of the target power supply vehicle, the navigation information of the target power supply path, and the setting information of the target power supply speed.

[0158] In some embodiments, the emergency monitoring and management platform can integrate the identification information of these target power supply vehicles, the navigation information of the target power supply path, and the setting information of the target power supply speed into the instruction as a second dispatch instruction, based on the target power supply strategy.

[0159] Step 540: Based on the second scheduling instruction, control the target power supply vehicle to move to the base station associated with the power outage area according to the target power supply path, and control the generator of the target power supply vehicle to supply power at the target power supply speed.

[0160] In some embodiments, the emergency monitoring and management platform can send the navigation information of the corresponding target power supply path and the setting information of the target power supply speed to the corresponding target power supply vehicle based on the identification information of the target power supply vehicle in the second dispatch instruction, thereby controlling the target power supply vehicle to move to the base station in the corresponding power outage area, so that the generator can supply power to the corresponding base station at the target power supply speed.

[0161] In some embodiments of this specification, by introducing a path optimization algorithm to determine the target power supply strategy, it is possible to comprehensively consider priority, buffer period and real-time road conditions to achieve a globally optimal scheduling strategy, which significantly improves the efficiency and success rate of multi-region and multi-vehicle collaborative emergency power supply.

[0162] The above descriptions of processes 200 and 500 are for illustrative purposes only and not for limitation. Those skilled in the art can make various modifications and changes to processes 200 and 500 under the guidance of this specification. These modifications and changes remain within the scope of this specification.

[0163] Some embodiments of this specification also provide a computer-readable storage medium that stores computer instructions. When a computer reads the computer instructions from the storage medium, the computer executes the above-described emergency power supply monitoring method for smart city base stations.

[0164] The basic concepts have been described above. Obviously, for those skilled in the art, the detailed disclosure above is merely illustrative and does not constitute a limitation of this specification. Those skilled in the art may make various modifications, improvements, and corrections to this specification. Such modifications, improvements, and corrections still fall within the spirit and scope of the exemplary embodiments in this specification.

Claims

1. A large-scale IoT model system for emergency monitoring of base station power supplies in smart cities, characterized in that, This includes an emergency monitoring and management platform, which is configured as follows: In response to the detection of a power outage in a region, at least one power outage region is identified; For each of the at least one power outage area: Based on the base station data of the power outage area, determine the backup power supply for the power outage area; Based on the power consumption characteristics, historical power consumption sequence, and historical state characteristics of the emergency events within the power outage area, a predicted power consumption sequence is determined using a power consumption prediction model. The power consumption prediction model is a machine learning model. The predicted power consumption sequence includes the predicted power consumption of the emergency events at multiple moments within a future period. The time interval between the multiple moments in the predicted power consumption sequence is related to the length of the confidence interval for the estimated power supply time. The historical state characteristics include historical power efficiency, historical ambient temperature, and historical equipment load rate. The estimated battery life of the power outage area is determined based on the reserve power, the base power consumption of the power outage area, and the predicted power consumption sequence. Based on the estimated battery life and estimated power supply time, a first scheduling instruction including the power supply time is generated and sent to the emergency monitoring and sensing control platform; as well as Based on the first scheduling instruction, the autonomous driving power supply vehicle is controlled to move to the base station associated with the power outage area at the power supply time, and the generator of the autonomous driving power supply vehicle is controlled to provide temporary power to the base station. The emergency monitoring and management platform is further configured as follows: Based on the time interval, a monitoring and control instruction is generated and sent to the emergency monitoring and control platform; Based on the monitoring and control instructions, the IoT sensors in the power outage area are controlled to collect data at a sampling rate and upload the collected data at an upload frequency.

2. The system according to claim 1, characterized in that, The emergency monitoring and management platform is further configured as follows: Power supply maps are constructed based on base station data and location relationships across multiple regions; Based on the power supply map, the confidence interval of the estimated power supply time is determined by a power supply prediction model, wherein the power supply prediction model is a machine learning model. The estimated power supply time is determined based on the confidence interval of the estimated power supply time.

3. The system according to claim 2, characterized in that, The power supply map includes a power outage feature extraction layer and a power outage time estimation layer; the input of the power outage feature extraction layer includes the power supply map, and the output includes power supply map features; the input of the power outage time estimation layer includes the power supply map features and electrical fault features, and the output includes the confidence interval of the estimated power supply time; the power outage feature extraction layer is a graph neural network model, and the power outage time estimation layer is a deep neural network model.

4. The system according to claim 1, characterized in that, The emergency monitoring and management platform is further configured as follows: The power supply priority of the at least one power outage area is determined based on the regional importance of the at least one power outage area, the level of event impact, and the urgency of battery life. Based on the power supply time, power supply priority, and location data of the autonomous power supply vehicle in the at least one power outage area, a target power supply strategy for the at least one power outage area is determined. The target power supply strategy includes a target power supply vehicle, a target power supply path, and a target power supply rotation speed. Based on the target power supply strategy, a second scheduling instruction is generated and sent to the emergency monitoring and sensing control platform; Based on the second scheduling command, the target power supply vehicle is controlled to move to the base station associated with the power outage area according to the target power supply path, and the generator of the target power supply vehicle is controlled to supply power at the target power supply speed.

5. A method for emergency monitoring of power supply for smart city base stations, characterized in that, The method is executed by the emergency monitoring and management platform of the smart city base station power emergency monitoring system, and includes: In response to the detection of a power outage in a region, at least one power outage region is identified; For each of the at least one power outage area: Based on the base station data of the power outage area, determine the backup power supply for the power outage area; Based on the power consumption characteristics, historical power consumption sequence, and historical state characteristics of the emergency events within the power outage area, a predicted power consumption sequence is determined using a power consumption prediction model. The power consumption prediction model is a machine learning model. The predicted power consumption sequence includes the predicted power consumption of the emergency events at multiple moments within a future period. The time interval between the multiple moments in the predicted power consumption sequence is related to the length of the confidence interval for the estimated power supply time. The historical state characteristics include historical power efficiency, historical ambient temperature, and historical equipment load rate. The estimated battery life of the power outage area is determined based on the reserve power, the base power consumption of the power outage area, and the predicted power consumption sequence. Based on the estimated battery life and estimated power supply time, a first scheduling instruction including the power supply time is generated and sent to the emergency monitoring and sensing control platform; and Based on the first scheduling instruction, the autonomous driving power supply vehicle is controlled to move to the base station associated with the power outage area at the power supply time, and the generator of the autonomous driving power supply vehicle is controlled to provide temporary power to the base station. The method further includes: Based on the time interval, a monitoring and control instruction is generated and sent to the emergency monitoring and control platform; Based on the monitoring and control instructions, the IoT sensors in the power outage area are controlled to collect data at a sampling rate and upload the collected data at an upload frequency.

6. The method according to claim 5, characterized in that, The method further includes: Power supply maps are constructed based on base station data and location relationships across multiple regions; Based on the power supply map, the confidence interval of the estimated power supply time is determined by a power supply prediction model, wherein the power supply prediction model is a machine learning model. The estimated power supply time is determined based on the confidence interval of the estimated power supply time.

7. The method according to claim 6, characterized in that, The power supply map includes a power outage feature extraction layer and a power outage time estimation layer; the input of the power outage feature extraction layer includes the power supply map, and the output includes power supply map features; the input of the power outage time estimation layer includes the power supply map features and electrical fault features, and the output includes the confidence interval of the estimated power supply time; the power outage feature extraction layer is a graph neural network model, and the power outage time estimation layer is a deep neural network model.

8. The method according to claim 5, characterized in that, The method further includes: The power supply priority of the at least one power outage area is determined based on the regional importance of the at least one power outage area, the level of event impact, and the urgency of battery life. Based on the power supply time, power supply priority, and location data of the autonomous power supply vehicle in the at least one power outage area, a target power supply strategy for the at least one power outage area is determined. The target power supply strategy includes a target power supply vehicle, a target power supply path, and a target power supply rotation speed. Based on the target power supply strategy, a second scheduling instruction is generated and sent to the emergency monitoring and sensing control platform; Based on the second scheduling command, the target power supply vehicle is controlled to move to the base station associated with the power outage area according to the target power supply path, and the generator of the target power supply vehicle is controlled to supply power at the target power supply speed.

Images