Intelligent zero-carbon-emission building energy-saving control method and system and medium

Abstract

The invention discloses an intelligent zero-carbon-emission building energy-saving control method and system and a medium, relates to the technical field of energy-saving control, and solves the technical problem of excessive energy supply or insufficient energy supply caused by the adoption of a one-step control strategy without distinguishing dynamic scenes such as conventional operation, mode switching and emergency situations. According to the method, a multi-source sensor data + hierarchical energy consumption calculation mode is adopted, a dynamic energy consumption baseline is combined, refined monitoring from total energy consumption to equipment level and partition level is achieved, the anomaly recognition accuracy is improved, building scenes are subdivided into conventional scenes, transitional scenes and sudden scenes based on a DBSCAN algorithm, the conventional scenes predict demands through machine learning and are cooperatively scheduled with equipment, and the probability that the equipment is abnormal is lowered. Energy supply according to needs and energy consumption reduction are achieved, transition scenes are controlled in a three-stage hierarchical mode, the energy consumption peak value caused by sudden starting and sudden stopping of equipment is avoided, emergency scenes are subjected to 10-second-level rapid recognition and priority energy distribution, and the emergency energy consumption waste is reduced while core functions are guaranteed.

Description

Technical Field

[0001] This invention relates to the field of energy-saving control technology, specifically to an intelligent zero-carbon emission building energy-saving control method, system, and medium. Background Technology

[0002] With the advancement of the global low-carbon transition, buildings, as a key area of ​​energy consumption and carbon emissions, have received widespread attention for their energy-saving control technologies.

[0003] According to patent application CN118963232A, an IoT-based building energy-saving control system is disclosed, including a regional control and early warning analysis module, a regional lighting alarm module, an energy-saving control execution module, a lighting equipment monitoring module, and a building monitoring terminal. This invention analyzes the current environmental conditions of corresponding corridor areas within a building through the regional control and early warning analysis module, and determines the lighting status of corresponding corridor areas through the regional lighting alarm module. This enables automatic and timely adjustment of the environment in each corridor area and automatic control of the lighting in each corridor area, significantly improving the building's energy-saving effect. Furthermore, the lighting equipment monitoring module analyzes the quality status of lighting equipment in each corridor area to enable timely replacement of lighting equipment, ensuring lighting effects while reducing energy consumption. This reduces the management difficulty for managers and demonstrates a high degree of intelligence.

[0004] However, some existing energy-saving control methods have a single dimension of energy consumption monitoring, focusing mainly on total energy consumption statistics. They lack a detailed breakdown of equipment types and functional areas, making it difficult to locate the root cause of abnormal energy consumption. The scenario identification is crude, failing to distinguish between dynamic scenarios such as normal operation, mode switching, and emergencies. They adopt a "one-size-fits-all" control strategy, resulting in over-supply or under-supply of energy. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an intelligent zero-carbon emission building energy-saving control method, system, and medium, which solves the problem of over- or under-energy supply caused by the use of a one-size-fits-all control strategy that fails to distinguish between dynamic scenarios such as normal operation, mode switching, and emergencies.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an intelligent zero-carbon emission building energy-saving control method, which specifically includes the following steps:

[0007] Collect multi-source energy consumption data and equipment energy consumption of buildings, combine historical data to predict short-term energy consumption, and generate abnormal or normal energy consumption signals by comparison.

[0008] Abnormal signals are processed in a graded manner according to their level and scope of impact. Key information is extracted and invalid signals are removed. A dynamic baseline is generated by combining historical data in dimensions such as time and scenario to locate the cause of the anomaly and generate anomaly information.

[0009] The scenarios corresponding to normal signals are classified into three categories—normal, transitional, and bursty—using the DBSCAN algorithm after feature processing.

[0010] For routine scenarios, machine learning is used to predict demand, optimize equipment parameters, combine personnel density control, and leverage photovoltaic and waste heat to generate routine energy-saving information.

[0011] The transition scenario uses a three-stage control based on the predicted load curve, combined with real-time deviation correction parameters, to generate transition energy-saving information.

[0012] In case of emergencies, sensors identify the type of emergency, allocate energy according to priority, coordinate with energy storage systems, and generate emergency energy-saving information.

[0013] As a further aspect of the present invention, the specific method for generating abnormal or normal energy consumption signals is as follows:

[0014] The system collects power, current, and energy consumption data from multi-source sensors in the building, as well as energy consumption data from smart devices. It calculates the total energy consumption of the devices, builds an energy consumption prediction model based on historical data, obtains short-term energy consumption data, and compares the total energy consumption with the short-term energy consumption. If the former is larger, an energy consumption anomaly analysis signal is generated; otherwise, a normal energy consumption analysis signal is generated.

[0015] As a further aspect of the present invention, the specific method for classifying the scenarios into three categories—regular, transitional, and sudden—using the DBSCAN algorithm after feature processing is as follows:

[0016] In normal scenarios, the standard deviation of intra-cluster features is <0.1, and the duration is ≥30 minutes. In transitional scenarios, intra-cluster features exhibit linear changes, and the duration is 10-30 minutes. In sudden scenarios, intra-cluster features undergo abrupt changes, and the duration is <10 minutes.

[0017] As a further aspect of the present invention, the specific method for generating conventional energy-saving information for conventional scenarios is as follows:

[0018] By using machine learning to predict short-term energy demand, the parameters of equipment such as air conditioners are optimized to maintain efficient operation. Based on the population density, energy is supplied on demand. The system is linked with HVAC and lighting systems, and photovoltaic and waste heat are used to meet low-grade demand. Surplus electrical energy is stored, and routine energy-saving control information is generated.

[0019] As a further aspect of the present invention, the specific method for generating transitional energy-saving information is as follows:

[0020] By combining historical data with real-time variables such as outdoor temperature and humidity and personnel arrival rate, the load growth curve is predicted, the target load of each node equipment is determined, and control is carried out in three stages: preparation, growth and target attainment.

[0021] In the preparation phase, core equipment is started with low power; in the growth phase, power is increased as personnel arrive at their posts; and in the target phase, power is stabilized to 90% of the target load. By combining the deviation correction parameters between real-time sensor data and predicted values, transitional energy-saving control information is generated.

[0022] As a further aspect of the present invention, the specific method for generating sudden energy-saving information is as follows:

[0023] For sudden scenarios, the scenario type can be identified within 10 seconds through high-density sensors and edge computing nodes; the core functions and loads that can be reduced can be identified, and energy allocation can be adjusted based on the priority matrix to generate sudden energy-saving control information.

[0024] An intelligent zero-carbon emission building energy-saving control system, the system comprising:

[0025] The data acquisition and energy consumption calculation module is used to collect multi-source sensor data and equipment energy consumption, calculate the total equipment energy consumption, and generate energy consumption abnormal / normal analysis signals.

[0026] The abnormal signal processing module is used to classify energy consumption anomaly analysis signals, extract features, filter historical data, locate the cause of anomalies, and generate energy consumption anomaly information.

[0027] The scenario classification and adjustment control module is used to classify building scenarios into regular scenarios, transitional scenarios, or sudden scenarios based on real-time features using the DBSCAN algorithm. It is used to predict the demand of regular scenarios, optimize equipment parameters and linkage scheduling, and generate regular energy-saving control information. It is also used to predict the load curve of transitional scenarios, control equipment in stages and dynamically correct parameters to generate transitional energy-saving control information. Furthermore, it is used to identify sudden scenarios, allocate energy based on the priority matrix, and generate sudden energy-saving control information.

[0028] The control information display module is used to display the generated energy consumption anomaly information, normal, transitional and sudden energy-saving control information to the corresponding management personnel.

[0029] As a further embodiment of the present invention, the data acquisition and energy consumption calculation module includes smart meters, current sensors, power sensors deployed according to functional zones, as well as infrared sensors for counting personnel density and temperature and humidity sensors for environmental parameters.

[0030] A computer-readable storage medium storing a computer program, which, when executed by a processor, implements an intelligent zero-carbon emission building energy-saving control method.

[0031] This invention provides an intelligent zero-carbon emission building energy-saving control method, system, and medium. Compared with existing technologies, it has the following advantages:

[0032] This invention achieves refined monitoring from total energy consumption to equipment level and zone level by using multi-source sensor data and a hierarchical energy consumption calculation mode, combined with a dynamic energy consumption baseline, thereby improving the accuracy of anomaly identification. At the same time, it automatically traces the cause of anomalies by matching historical data in four dimensions, reducing the false alarm and missed alarm rates.

[0033] This invention uses the DBSCAN algorithm to subdivide building scenarios into three categories: normal, transitional, and emergency. In normal scenarios, machine learning is used to predict demand and coordinate equipment scheduling to achieve on-demand energy supply and reduce energy consumption. In transitional scenarios, three-stage tiered control is used to avoid energy consumption peaks caused by sudden equipment start-up and shutdown. In emergency scenarios, rapid identification and priority energy allocation within 10 seconds are used to reduce emergency energy waste while ensuring core functions. Attached Figure Description

[0034] Figure 1 This is a diagram illustrating the steps and methods of the present invention;

[0035] Figure 2 This is the system block diagram of the present invention. Detailed Implementation

[0036] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0037] Example 1

[0038] Please see Figure 1 This application provides an intelligent zero-carbon emission building energy-saving control method, which specifically includes the following steps:

[0039] Step 1: Based on the multi-source sensors installed in the building, acquire their corresponding multi-source sensor data, including power, current, and energy consumption. Simultaneously, acquire the energy consumption of the smart devices within the building and calculate the total energy consumption of the building's devices. The total energy consumption is broken down into two levels: "functional zone - device type," forming a traceable energy consumption structure. Level 1 (functional zone): Office area, meeting room, public corridor, equipment room, underground parking garage, etc., calculate the energy consumption percentage for each zone (e.g., a sudden increase in the energy consumption percentage of the equipment room may indicate a water pump malfunction). Level 2 (equipment type): Within each zone, statistics are compiled according to "air conditioning system (including cold and heat sources, terminal equipment), lighting system, power equipment (elevators, water pumps), other electricity consumption (sockets, charging piles)." For example: Total energy consumption = Σ(air conditioning energy consumption of each zone + lighting energy consumption + power equipment energy consumption + other energy consumption).

[0040] Next, a corresponding energy consumption prediction model is established based on historical data, and the corresponding short-term energy consumption is obtained. The total equipment energy consumption is then compared with the short-term energy consumption.

[0041] If the total energy consumption of the equipment is greater than the short-term energy consumption, it indicates that the overall energy consumption of the building is abnormal, and an energy consumption abnormality analysis signal is generated. Conversely, if the equipment energy consumption is less than the short-term energy consumption, it indicates that the overall energy consumption of the building is normal, and an energy consumption normality analysis signal is generated.

[0042] Step 2: Process the acquired energy consumption anomaly analysis signals, obtain historical data, compare the obtained real-time data with the historical data, identify the corresponding anomaly data, and combine the sensing data to locate the anomaly, determine the cause of the anomaly, and generate energy consumption anomaly information.

[0043] Based on the anomaly level ("warning interval / anomaly interval") and the scope of impact (single device / single area / entire building), it is divided into three levels:

[0044] Level 1 Signal (Minor Warning): The energy consumption of a single device slightly exceeds the threshold (e.g., the energy consumption of an office socket is 10%-15% higher than the historical average), with no significant impact. The processing time requirement is ≤24 hours.

[0045] Level 2 signal (area anomaly): Multiple devices in a single area have abnormal energy consumption (e.g., the total energy consumption of the 12-story office area exceeds the predicted value by 15%-20%), which may affect local operation. The processing time requirement is ≤2 hours.

[0046] Level 3 signal (serious anomaly): Energy consumption of cross-regional or core equipment (such as central air conditioning unit) suddenly increases (more than 20% above the predicted value), which may cause safety risks (such as equipment overload). The processing time requirement is ≤15 minutes.

[0047] Upon receiving a signal, key features are immediately extracted, including the automatic parsing of the core information contained in the signal: the time of the anomaly, the area / equipment involved, real-time energy consumption data (e.g., current power 12kW, historical average 8kW), and associated sensor alarms (e.g., "air conditioner compressor current exceeds limit").

[0048] Eliminate invalid signals: Filter out false alarms caused by sensor malfunctions (such as data jumps) and data transmission errors (such as confirming that an abnormal meter reading is due to equipment failure through cross-verification of sensor data from adjacent areas).

[0049] Comparable historical data was filtered based on four dimensions: time, scenario, device status, and environment.

[0050] Time dimension: Prioritize matching “data from the same period” (e.g., for an anomaly at 9:00 this Wednesday, match data from 9:00 Wednesdays ± 30 minutes in the past 4 weeks). If the data volume is insufficient, expand to the same week and time period (e.g., 9:00 on other weekdays).

[0051] Scenario Dimension: Ensure that historical data is consistent with the current scenario (e.g., both are "regular office scenarios", excluding special scenarios such as holidays and ad-hoc meetings).

[0052] Equipment status dimension: Filter historical periods where equipment has no faults and no maintenance records (e.g., exclude abnormal data within 1 week after an air conditioner was repaired).

[0053] Environmental dimension: Match historical data with similar outdoor temperature, humidity, light intensity and other environmental parameters (e.g., if the current outdoor temperature is 28℃, prioritize historical data of 27-29℃ for the same period).

[0054] Statistical analysis is performed on the matched historical data to generate a "dynamic energy consumption baseline" instead of a fixed value:

[0055] Calculate the mean, standard deviation, and 95% confidence interval of historical data (e.g., mean 5 kWh, standard deviation 0.8 kWh, confidence interval 3.4-6.6 kWh) as the benchmark for judging anomalies (current real-time energy consumption of 8 kWh exceeds the upper limit of the confidence interval and is confirmed as abnormal data).

[0056] For scenarios with large fluctuations (such as transitional scenarios), the baseline must include "trend characteristics" (such as energy consumption should increase linearly from 8:00 to 9:00 in the morning with a slope of 0.5 kWh / 10 minutes). If the real-time data deviates from the trend (such as the slope suddenly changing to 1.2 kWh / 10 minutes), it is judged as abnormal.

[0057] Based on historical cases, common causal relationships are identified, forming a rule base for rapid matching:

[0058] If the air conditioner has "high energy consumption + high compressor frequency + indoor temperature lower than the set value" → possible causes: temperature control sensor malfunction, return air valve stuck.

[0059] If the message is “high lighting energy consumption + 0 people density + light switch status is on”, possible causes are: user forgot to turn off the light or smart switch malfunction.

[0060] If the elevator has "high energy consumption + normal number of runs + high motor temperature", the possible causes are: insufficient lubrication of the guide rails or aging of the motor.

[0061] Step 3: Process the generated energy consumption normal analysis signal to obtain the current building scene, and classify it into normal scene, transitional scene, and sudden scene. The specific classification method is as follows:

[0062] Real-time building features are acquired, including time features, personnel features, energy consumption features, and environmental and equipment features. These features are then normalized and outlier filtered. The normalization process maps all feature values ​​to the [0, 1] interval (e.g., personnel density = actual number of people / maximum capacity of the area; energy consumption normalization = current energy consumption / historical maximum energy consumption), preventing any single feature (e.g., absolute energy consumption) from dominating the clustering results. Outlier filtering uses the 3σ criterion to remove extreme values ​​caused by sensor malfunctions (e.g., personnel density = 2, far exceeding the physical limit of 1), retaining real fluctuation data (e.g., personnel density = 1.2 in a sudden scenario due to temporary overcrowding). Data is then collected in units of time t, and a window feature vector is generated using the mean / trend value of the data within the window (e.g., the slope of personnel density change). Real-time classification is then performed using DBSCAN.

[0063] Typical scenario: Intra-cluster feature standard deviation <0.1 (e.g., personnel density fluctuation ±5%, energy consumption fluctuation ±8%), duration ≥30 minutes → output typical scenario.

[0064] Transitional scenario: Cluster features change linearly (e.g., population density increases by 0.02 per minute, energy consumption increases by 0.03 per minute), lasting 10-30 minutes → Output transitional scenario.

[0065] Sudden Situation: Abrupt changes in cluster characteristics (e.g., personnel density increases by 0.5 and energy consumption increases by 0.4 within 5 minutes), lasting less than 10 minutes → Output sudden scenario.

[0066] For example, an office building uses a sampling frequency of 1 minute / sample and a 5-minute window to generate feature vectors, which are then classified in real time using DBSCAN.

[0067] 8:00-8:30: The window features show that "personnel density increases from 0.1 to 0.5, energy consumption increases from 0.2 to 0.6, and air conditioning frequency increases from 30Hz to 50Hz". The features continue to change (transition features) and are clustered into cluster 2 → output "transition scene".

[0068] 10:00-16:00: Window features show "personnel density is stable at 0.6-0.7, energy consumption is 0.7-0.8, and air conditioning frequency is stable at 50Hz", feature fluctuation <5% → classified into cluster 1 → output "normal scene".

[0069] 14:00-14:10: 20 people suddenly flood into a conference room (personnel density suddenly increases from 0.2 to 1.2), energy consumption increases from 0.3 to 1.0 within 5 minutes → clustered as noise points, manually labeled as "sudden meeting scenario" → output "sudden scenario".

[0070] Step 4: Implement energy-saving control for the classified routine scenarios. Use machine learning models (such as LSTM and random forest) to analyze historical energy consumption data, personnel flow patterns, and meteorological data to predict the demand (such as temperature, lighting brightness, and fresh air volume) for the next 4-24 hours.

[0071] For example: If historical data shows that the occupancy rate in an office is consistently around 80% from 10:00 AM to 3:00 PM on weekdays, then the air conditioning temperature setting should be adjusted in advance (from 26℃ to 27℃ in summer, and from 20℃ to 19℃ in winter), and a fresh air volume of 30m³ should be matched accordingly. 3 / h·person), avoid operating at "full load";

[0072] Optimize equipment operating parameters, such as the outlet water temperature and fan speed of air conditioners, and maintain the COP (coefficient of performance) in the optimal range through dynamic adjustment (e.g., COP ≥ 5.0 for centrifugal chillers). Combine this with real-time monitoring of personnel density using infrared sensors, cameras, or WiFi probes to achieve "energy supply on demand when people are present, and rapid load reduction when people leave."

[0073] For example: when the meeting room is empty, the lights and air conditioning will be automatically turned off within 5 minutes (while maintaining minimum ventilation); when the personnel density in the open office area is less than 30%, 1 / 3 of the lighting circuits will be automatically turned off, and the air conditioning temperature will be adjusted up or down by 1-2℃.

[0074] Equipment coordinated scheduling: For example, when the HVAC system is linked with the lighting system, when there is sufficient natural light (illuminance ≥ 500 lux), the brightness of the lights in the corresponding area will be automatically reduced (from 80% to 50%), and the return air temperature of the air conditioner will be adjusted at the same time (using the heat from natural light to reduce the cooling load).

[0075] In conventional scenarios, photovoltaic power generation and waste heat (such as air conditioning condensation heat) are prioritized for supplying low-grade needs (such as hot water supply and underground garage lighting), and the remaining electrical energy is stored in energy storage batteries to avoid direct grid connection and waste, and generates conventional energy-saving control information.

[0076] For example, photovoltaic panels in office buildings generate electricity during the day to power lighting, and excess electricity is stored in batteries and used for elevator operation at night.

[0077] Step 5: Implement energy-saving control for the transition scenarios obtained from the classification. Combine historical data (such as the load curve of the transition period in the past 3 months) with real-time variables (such as the weather on the day and personnel appointment information) to predict the load growth curve of the transition scenarios (such as the air conditioning load needs to gradually increase from 10kW to 50kW within 1 hour from "non-working mode" to "working mode", rather than jumping instantaneously).

[0078] Input variables: outdoor temperature and humidity (affecting air conditioning load), staff arrival rate (predicted through the attendance system), and equipment preheating time (e.g., it takes 40 minutes for the air conditioner to go from shutdown to stable operation).

[0079] Output target: The target load of equipment at each time point (e.g., 15kW air conditioning load from 7:00 to 7:10, increasing to 25kW from 7:10 to 7:20) to ensure that the target is met just when personnel enter, rather than over-supplying energy in advance.

[0080] The transition scenario is divided into three stages: "preparation, growth, and achievement," with differentiated equipment operation goals set for each stage to avoid "achieving everything in one step."

[0081] Preparation phase (e.g., 7:00-7:20 in the morning): Only start the low-power mode of the core equipment (e.g., the air conditioner runs at 30% power, and the target temperature is 2°C lower than the final value) to meet the needs of a small number of early arrivals, and keep energy consumption within 30% of the target load.

[0082] During the growth phase (e.g., 7:20-7:50): increase equipment power as personnel arrive at their posts (increase by 10%-15% every 10 minutes), gradually approach the set value of the target temperature (e.g., from 22℃ to 24℃), and control energy consumption at 50%-80% of the target load.

[0083] During the target achievement phase (e.g., 7:50-8:00): the equipment power is stabilized at 90% of the target load (with a 10% buffer to cope with fluctuations), and parameters such as temperature / brightness are accurately met (e.g., air conditioner 26℃±0.5℃).

[0084] Simultaneously, data from sensors deployed in different areas is acquired, and their real-time status is collected. The deviation is then calculated by comparing the data with the predicted values.

[0085] If the actual personnel density is lower than predicted (e.g., the attendance rate on a certain floor is only 50%), immediately reduce the equipment load in that area (e.g., reduce the air conditioning power from 30kW to 20kW).

[0086] If the outdoor temperature drops suddenly (e.g., the temperature on a winter morning is 5°C lower than predicted), increase the air conditioner's heating rate appropriately (from 1°C every 10 minutes to 1.2°C) to avoid delays in reaching the target temperature.

[0087] Based on the deviation, the equipment operating parameters are dynamically corrected to generate transitional energy-saving control information.

[0088] Step Six: Implement energy-saving control for the identified emergency scenarios, deploy high-density sensors and edge computing nodes, and complete data acquisition and scenario assessment within 10 seconds.

[0089] Emergency personnel scenarios: Real-time statistics on the density of people in the area are collected using infrared sensors / cameras (e.g., the number of people in a meeting room increases from 5 to 30 within 5 minutes), and access control data is used to determine whether they are temporary visitors (e.g., for external meetings).

[0090] Equipment failure scenarios: Identify the type of failure (such as air conditioner shutdown or water pump jamming) by using equipment status sensors (such as vibration value of air conditioner compressor or sudden change in elevator current) and simultaneously locate the affected area (such as all air conditioners on a certain floor stopping).

[0091] Environmental emergencies: Access weather warnings (such as a sudden temperature rise of 8°C within 1 hour) and power grid signals (such as voltage drops and power outage warnings) to predict the impact on building load (such as extreme high temperatures causing a 30% surge in air conditioning load).

[0092] Clearly define the "core functions that must be guaranteed" and the "non-core loads that can be temporarily reduced" for different scenarios to avoid indiscriminate power supply:

[0093] For sudden outbreaks involving large crowds (such as temporary exhibitions): the core requirements are "ventilation (CO2≤1000ppm) + basic cooling (room temperature≤28℃)", which can reduce lighting in non-exhibition areas and unnecessary equipment (such as idle office air conditioners).

[0094] Sudden equipment failure (such as air conditioning unit shutdown): The core requirement is "temperature control in critical areas (such as operating rooms and server rooms)," which can sacrifice the comfort of public areas (such as relaxing the corridor temperature to 30°C).

[0095] In the event of a power grid emergency (such as power rationing), the core demand is for "emergency lighting + power supply for critical equipment (such as fire protection systems)," requiring a rapid switch to energy storage or backup power sources and the shutdown of all non-essential loads.

[0096] Based on a priority matrix, energy distribution can be quickly adjusted using intelligent circuit breakers, valves, and other actuators.

[0097] In the event of a sudden surge in personnel (such as a temporary meeting of 30 people): The system automatically reduces the power consumption of air conditioning in adjacent offices by 20% and halves the lighting in public corridors (level 3 load), transferring the released cooling and electricity to the meeting room (increasing its fresh air volume to 40m³ / h). 3 / h·person, to ensure CO2 does not exceed the standard).

[0098] Equipment failure (e.g., air conditioning shutdown on a certain floor): The redundant cooling capacity of the secondary load air conditioning on other floors (e.g., the computer room air conditioning set temperature is increased from 22℃ to 23℃, releasing 10% of the cooling capacity) is transferred to the faulty floor through pipe valves to avoid starting the high-energy-consuming standby generator.

[0099] Based on the above analysis, emergency energy-saving control information is generated.

[0100] Example 2

[0101] Please see Figure 2 This application provides an intelligent zero-carbon emission building energy-saving control system, which includes:

[0102] The data acquisition and energy consumption calculation module is used to collect multi-source sensor data and equipment energy consumption, calculate the total equipment energy consumption and generate energy consumption abnormal / normal analysis signals, and the specific processing method is the same as the processing process in step one.

[0103] The abnormal signal processing module is used to classify energy consumption anomaly analysis signals, extract features, filter historical data and locate the cause of anomalies, and generate energy consumption anomaly information. The specific processing method is the same as the processing process in step two.

[0104] The scene classification and adjustment control module is used to classify building scenes into normal scenes, transitional scenes, or sudden scenes based on real-time features using the DBSCAN algorithm. The specific processing method is the same as that in step three. It is used to predict the demand of normal scenes, optimize equipment parameters and linkage scheduling, and generate normal energy-saving control information. The specific processing method is the same as that in step four. It is also used to predict the load curve of transitional scenes, control equipment in stages and dynamically correct parameters to generate transitional energy-saving control information. The specific processing method is the same as that in step five. It is also used to identify sudden scenes, allocate energy based on the priority matrix, and generate sudden energy-saving control information. The specific processing method is the same as that in step six.

[0105] The control information display module is used to display the generated energy consumption anomaly information, normal, transitional and sudden energy-saving control information to the corresponding management personnel.

[0106] Example 3

[0107] Another embodiment of this application provides a computer storage medium for storing a computer program, which, when executed, is used to implement an intelligent zero-carbon emission building energy-saving control method.

[0108] Computer storage media, including both permanent and non-permanent, removable and non-removable media, can store information using any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0109] The data in the above formulas are all calculated using numerical values, without substituting the units of the parameters. In addition, the contents not described in detail in this specification are all prior art known to those skilled in the art.

[0110] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. An intelligent zero-carbon emission building energy-saving control method, characterized in that, The method specifically includes the following steps: Collect multi-source energy consumption data and equipment energy consumption of buildings, combine historical data to predict short-term energy consumption, and generate abnormal or normal energy consumption signals by comparison. Abnormal signals are processed in a graded manner according to their level and scope of impact. Key information is extracted and invalid signals are removed. A dynamic baseline is generated by combining historical data in dimensions such as time and scenario to locate the cause of the anomaly and generate anomaly information. The scenarios corresponding to normal signals are classified into three categories—normal, transitional, and bursty—using the DBSCAN algorithm after feature processing. For routine scenarios, machine learning is used to predict demand, optimize equipment parameters, combine personnel density control, and leverage photovoltaic and waste heat to generate routine energy-saving information. The transition scenario uses a three-stage control based on the predicted load curve, combined with real-time deviation correction parameters, to generate transition energy-saving information. In case of emergencies, sensors identify the type of emergency, allocate energy according to priority, coordinate with energy storage systems, and generate emergency energy-saving information.

2. The intelligent zero-carbon emission building energy-saving control method according to claim 1, characterized in that, The specific method for generating abnormal or normal energy consumption signals is as follows: The system collects power, current, and energy consumption data from multi-source sensors in the building, as well as energy consumption data from smart devices. It calculates the total energy consumption of the devices, builds an energy consumption prediction model based on historical data, obtains short-term energy consumption data, and compares the total energy consumption with the short-term energy consumption. If the former is larger, an energy consumption anomaly analysis signal is generated; otherwise, a normal energy consumption analysis signal is generated.

3. The intelligent zero-carbon emission building energy-saving control method according to claim 1, characterized in that, The specific method for classifying scenarios into three categories—regular, transitional, and sudden—using the DBSCAN algorithm after feature processing is as follows: In normal scenarios, the standard deviation of intra-cluster features is <0.1, and the duration is ≥30 minutes. In transitional scenarios, intra-cluster features exhibit linear changes, and the duration is 10-30 minutes. In sudden scenarios, intra-cluster features undergo abrupt changes, and the duration is <10 minutes.

4. The intelligent zero-carbon emission building energy-saving control method according to claim 1, characterized in that, The specific method for generating conventional energy-saving information for conventional scenarios is as follows: By using machine learning to predict short-term energy demand, the parameters of equipment such as air conditioners are optimized to maintain efficient operation. Based on the population density, energy is supplied on demand. The system is linked with HVAC and lighting systems, and photovoltaic and waste heat are used to meet low-grade demand. Surplus electrical energy is stored, and routine energy-saving control information is generated.

5. The intelligent zero-carbon emission building energy-saving control method according to claim 1, characterized in that, The specific method for generating transitional energy-saving information is as follows: By combining historical data with real-time variables such as outdoor temperature and humidity and personnel arrival rate, the load growth curve is predicted, the target load of each node equipment is determined, and control is carried out in three stages: preparation, growth and target attainment. In the preparation phase, core equipment is started with low power; in the growth phase, power is increased as personnel arrive at their posts; and in the target phase, power is stabilized to 90% of the target load. By combining the deviation correction parameters between real-time sensor data and predicted values, transitional energy-saving control information is generated.

6. The intelligent zero-carbon emission building energy-saving control method according to claim 1, characterized in that, The specific method for generating sudden energy-saving information is as follows: For sudden scenarios, the scenario type can be identified within 10 seconds through high-density sensors and edge computing nodes; the core functions and loads that can be reduced can be identified, and energy allocation can be adjusted based on the priority matrix to generate sudden energy-saving control information.

7. An intelligent zero-carbon emission building energy-saving control system, used to execute the intelligent zero-carbon emission building energy-saving control method according to any one of claims 1-6, characterized in that, include: The data acquisition and energy consumption calculation module is used to collect multi-source sensor data and equipment energy consumption, calculate the total equipment energy consumption, and generate energy consumption abnormal / normal analysis signals. The abnormal signal processing module is used to classify energy consumption anomaly analysis signals, extract features, filter historical data, locate the cause of anomalies, and generate energy consumption anomaly information. The scenario classification and adjustment control module is used to classify building scenarios into regular scenarios, transitional scenarios, or sudden scenarios based on real-time features using the DBSCAN algorithm. It is used to predict the demand of regular scenarios, optimize equipment parameters and linkage scheduling, and generate regular energy-saving control information. It is also used to predict the load curve of transitional scenarios, control equipment in stages and dynamically correct parameters to generate transitional energy-saving control information. Furthermore, it is used to identify sudden scenarios, allocate energy based on the priority matrix, and generate sudden energy-saving control information. The control information display module is used to display the generated energy consumption anomaly information, normal, transitional and sudden energy-saving control information to the corresponding management personnel.

8. The intelligent zero-carbon emission building energy-saving control system according to claim 7, characterized in that, The data acquisition and energy consumption calculation module includes smart meters, current sensors, power sensors deployed according to functional zones, as well as infrared sensors for counting personnel density and temperature and humidity sensors for environmental parameters.

9. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the intelligent zero-carbon emission building energy-saving control method as described in any one of claims 1-6.

Images