Multi-dimensional energy-saving analysis dynamic management platform and method for building

Abstract

The invention discloses a multi-dimensional energy-saving analysis dynamic management platform and method for a building, and relates to the technical field of building energy-saving management, and the method comprises the following steps: monitoring the indoor temperature in real time through a temperature monitoring system, and transmitting the collected temperature data to a control system; the obtained current indoor temperature is compared with a preset starting value and a preset stopping value of the air conditioning system, and starting and stopping of the air conditioning system are controlled through a control system; real-time indoor temperature data are collected and constructed into an analysis data set, and key features capable of reflecting the temperature change trend are extracted from the analysis data set. Through real-time temperature data analysis and machine learning prediction, temperature change stages are accurately divided, conventional start-stop control is adopted in a normal stage, output power is dynamically adjusted through fuzzy logic in a rapid stage, frequent start-stop and energy waste are avoided, temperature control precision and comfort are improved, energy efficiency is remarkably improved, operation cost is reduced, and the method is suitable for large-scale popularization and application. The service life of air conditioning equipment is prolonged, and an efficient and intelligent energy-saving management scheme is provided.

Description

Technical Field

[0001] The present invention relates to the technical field of building energy-saving management, and in particular to a multi-dimensional energy-saving analysis dynamic management platform and method for buildings. Background Art

[0002] The multi-dimensional energy-saving analysis dynamic management platform is an integrated system that aims to achieve comprehensive monitoring and optimized management of energy consumption through real-time data collection, analysis and prediction. The platform uses energy consumption data input from multiple dimensions, such as equipment usage status, environmental parameters, workload, energy consumption type, etc., combined with advanced data analysis algorithms and machine learning models, to dynamically evaluate various energy efficiency indicators and provide targeted energy-saving suggestions. The platform can monitor energy consumption trends in real time, identify potential waste areas, and adjust energy efficiency according to different time periods, working conditions or external environmental factors. Through flexible management strategies and feedback mechanisms, the platform helps enterprises achieve continuous energy optimization, reduce energy consumption costs, and improve energy efficiency. It has strong adaptability and scalability and can be customized according to different application scenarios and needs.

[0003] The prior art has the following deficiencies:

[0004] In the prior art, traditional air conditioners usually control the start and stop by setting the temperature. When the indoor temperature reaches the set stop value, the air conditioner stops working; when the temperature deviates from the set start value, the air conditioner restarts. Although this control method is simple and easy, when the temperature environment changes rapidly, the frequent start and stop of the air conditioner will cause indoor temperature fluctuations, making it difficult to maintain a stable and comfortable environment. At the same time, each time the air conditioner is started, the air conditioner compressor needs to consume a lot of energy to quickly adjust the temperature, and after stopping, the system enters a standby state, resulting in energy waste. Especially at startup, the compressor requires a large amount of power to overcome the initial load, resulting in low energy efficiency. Therefore, the traditional on / off control method is not efficient in some environments. Frequent start and stop not only affects comfort, but also increases unnecessary energy consumption.

[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0006] The purpose of the present invention is to provide a multi-dimensional energy-saving analysis dynamic management platform and method for buildings. Through the collection and analysis of real-time temperature data, combined with the prediction of a machine learning model, the temperature change stage can be accurately divided, so that conventional start-stop control can be adopted in the normal temperature change stage to reduce system complexity and energy consumption; in the rapid temperature change stage, the air-conditioning output power is dynamically adjusted through fuzzy logic to ensure the continuous operation of the air-conditioning, maintain room temperature stability and avoid energy waste and equipment wear caused by frequent start-up of the compressor. It not only improves the comfort of the indoor environment and the temperature control accuracy, but also significantly improves the energy efficiency of the system, reduces operating costs, and extends the service life of air-conditioning equipment, so as to solve the problems in the above-mentioned background technology.

[0007] In order to achieve the above object, the present invention provides the following technical solution: a multi-dimensional energy-saving analysis and dynamic management method for a building, comprising the following steps:

[0008] Monitor the indoor temperature in real time through the temperature monitoring system and transmit the collected temperature data to the control system;

[0009] The current indoor temperature is compared with the preset start value and stop value of the air conditioning system, and the start and stop of the air conditioning system are controlled by the control system;

[0010] Collect and construct real-time indoor temperature data into an analytical dataset, from which key features that can reflect temperature change trends are extracted;

[0011] Under the monitoring window, the extracted features are deeply analyzed and the processed data is input into the pre-trained machine learning model to identify and understand the law of indoor temperature changes and predict future temperature change trends;

[0012] The indoor temperature change is predicted through the machine learning model, and the indoor temperature change is divided into the normal temperature change stage and the rapid temperature change stage;

[0013] During the normal temperature change phase, the air conditioning system continues to monitor the current indoor temperature, compares it with the set start and stop values, and controls the start and stop of the air conditioning in the usual way;

[0014] During the rapid temperature change stage, fuzzy logic is used to control the output power of the air conditioner according to the indoor temperature changes, so that the current indoor temperature is always between the start value and the stop value, ensuring that the air conditioner always remains in operation, maintaining the indoor temperature while avoiding frequent start and stop of the air conditioner.

[0015] Preferably, the temperature monitoring system is a plurality of temperature sensors evenly arranged in the indoor area, and temperature data is collected through the plurality of temperature sensors, and the average value of the data collected by the plurality of temperature sensors is calculated to obtain the current indoor temperature.

[0016] Preferably, key features that can reflect the temperature change trend are extracted, wherein the extracted key features include the speed at which the indoor real-time temperature deviates from the set value and the frequency and intensity of the real-time temperature fluctuation. After the key features are obtained, the speed at which the indoor real-time temperature deviates from the set value and the frequency and intensity of the real-time temperature fluctuation are deeply analyzed under the monitoring window, and a temperature deviation speed quantization value and a temperature fluctuation quantization value are generated respectively. The temperature deviation speed quantization value reflects the rapidity of the temperature change; the temperature fluctuation quantization value reflects the volatility and amplitude of the temperature change;

[0017] After obtaining the temperature deviation rate quantification value and temperature fluctuation quantification value generated by in-depth analysis of the extracted key features, the temperature deviation rate quantification value and the temperature fluctuation quantification value are input into a pre-trained machine learning model to generate a dynamic temperature change coefficient, and the future temperature change trend is predicted through the dynamic temperature change coefficient.

[0018] Preferably, the dynamic temperature change coefficient generated when predicting the indoor temperature change trend under the monitoring window by the pre-trained machine learning model is compared and analyzed with the pre-set dynamic temperature change coefficient reference threshold, and the indoor temperature change is divided. The specific division steps are as follows:

[0019] If the dynamic temperature change coefficient generated when predicting the indoor temperature change trend through the pre-trained machine learning model under the monitoring window is greater than the dynamic temperature change coefficient reference threshold, the indoor temperature change is divided into a rapid temperature change stage;

[0020] If the dynamic temperature change coefficient generated when predicting the indoor temperature change trend through the pre-trained machine learning model under the monitoring window is less than or equal to the dynamic temperature change coefficient reference threshold, the indoor temperature change is classified as the normal temperature change stage.

[0021] Preferably, in the monitoring window, the specific steps of conducting an in-depth analysis of the speed at which the indoor real-time temperature deviates from the set value to generate a quantified value of the temperature deviation speed are as follows:

[0022] In the monitoring window, record the real-time indoor temperature at each time point and the target temperature set by the air conditioning system, and calculate the temperature deviation rate at each time point. The calculation expression is:

[0023]

[0024] , where T(t) is the real-time indoor temperature at time t, T s is the target temperature set for the air conditioner, R(t) is the temperature deviation rate at time t, expressed as a percentage;

[0025] At each time point, the rate of change of the temperature deviation rate relative to time, that is, the deviation rate acceleration, is calculated. The calculation expression is:

[0026]

[0027] , where A(t) is the deviation rate acceleration at time t, indicating the speed of change of temperature deviation rate, and Δt is the time interval between adjacent time points;

[0028] The temperature deviation velocity quantification value is calculated based on the deviation rate acceleration to quantify the rapidity of temperature change. The formula is:

[0029]

[0030] , where TDSI xy The quantized value of temperature deviation speed, Δ and β weighting factors, are used to control the sensitivity to deviation rate and temperature deviation rate, e -β·|R(t)| is an exponential decay term, t0 and t1 are the starting time point and the ending time point of the monitoring window respectively.

[0031] Preferably, in the monitoring window, the specific steps of conducting an in-depth analysis of the frequency and intensity of indoor real-time temperature fluctuations to generate a temperature fluctuation quantification value are as follows:

[0032] The time series feature extraction of indoor real-time temperature fluctuation data is performed. Assume that the temperature data of N time points are collected in the monitoring window, which is recorded as T i ={T1, T2, ..., T N}, where T i is the temperature value at the i-th time point;

[0033] Calculate the amplitude of temperature change at adjacent time points to obtain the temperature difference sequence. The calculation expression of the temperature difference sequence is:

[0034] ΔT i =T i+1 -T i

[0035] , where ΔT i represents the i-th value of the temperature difference sequence, which represents the temperature variation amplitude of two adjacent sampling points, i=1, 2, 3, ..., N-1;

[0036] A cumulative fluctuation energy function is defined to quantify the fluctuation intensity. It is combined with the sum of squares of the difference sequence and the time weight to measure the overall severity of the temperature fluctuation. The expression of the cumulative fluctuation energy function is:

[0037]

[0038] , where ET The accumulated fluctuation energy represents the intensity of temperature fluctuation within the monitoring window, w i is the time weight factor, which is used to highlight the contribution of fluctuations at different moments to the overall energy;

[0039] The temperature fluctuation quantification value is calculated by combining the accumulated fluctuation energy and the consistency of the temperature change direction to comprehensively reflect the amplitude of the fluctuation. The comprehensive calculation expression of the temperature fluctuation quantification value is:

[0040]

[0041] , where, TBI xy Indicates the quantitative value of temperature fluctuation, reflecting the volatility and amplitude of temperature change, sign(ΔT i ΔT i+1 ) indicates that the adjacent temperature difference (ΔT i ΔT i+1 ) is consistent to calculate the directional consistency of temperature change. δ is the directional consistency weight coefficient, which is used to balance the contribution of fluctuation intensity and directional consistency.

[0042] Preferably, in the rapid temperature change stage, fuzzy logic is used to control the output power of the air conditioner according to the indoor temperature change, so that the current indoor temperature is always between the start value and the stop value. The specific steps are as follows:

[0043] In the rapid temperature change stage, the dynamic temperature change coefficient is first compared with the dynamic temperature change coefficient reference threshold to determine the severity of the current temperature change, and as one of the input variables of the fuzzy logic, the relative position between the current indoor temperature and the start value and stop value of the air conditioner is calculated to generate the temperature deviation factor. The expression for generating the temperature deviation factor is:

[0044]

[0045] , where ΔT relative is the temperature deviation factor, which quantifies the degree of deviation of the current temperature within the target range. The temperature deviation factor ranges from [0, 1] and indicates the degree of deviation of the current temperature relative to the start value and the stop value. current is the current indoor temperature, T start is the starting value when the air conditioner is started, T stop It is the stop value when the air conditioner stops;

[0046] Calculate the normalized value of the dynamic temperature change amplitude. The calculation expression is:

[0047]

[0048] , among which, Temper dynaIndicates the dynamic temperature change coefficient, DTC_ref indicates the dynamic temperature change coefficient reference threshold, ΔT dynamic is a normalized value of dynamic temperature change, quantifying the ratio of the dynamic temperature change coefficient to a reference threshold value of the dynamic temperature change coefficient;

[0049] The temperature deviation factor ΔT relative and the normalized value of dynamic temperature change ΔT dynamic It will serve as the core input of the fuzzy logic system to analyze the drastic change of the current temperature and its position in the target range.

[0050] Preferably, the temperature deviation factor ΔT is based on the input relative and the normalized value of dynamic temperature change ΔT dynamic , the air conditioner output power adjustment coefficient is generated by fuzzy logic rules. The fuzzy logic rules consider two aspects:

[0051] When the temperature is close to the start value or stop value and the temperature changes dramatically, increase the output power to quickly stabilize the temperature;

[0052] When the temperature is far away from the start value and the stop value and the temperature changes smoothly, reduce the power to avoid excessive response;

[0053] The calculation expression of fuzzy logic output is:

[0054] P adjust =min(1,max(0,k p ΔT relative +k q ΔT dynamic ))

[0055] , where k p and k q are weight coefficients, which respectively control the influence of relative temperature deviation and dynamic change on output power adjustment, P adjust is the air conditioner output power adjustment coefficient, which is used to dynamically adjust the output power of the air conditioner. Min and max are used to limit the value range of the output power adjustment coefficient.

[0056] Preferably, the actual output power of the air conditioner is calculated according to the generated adjustment coefficient. At this stage, it is necessary to ensure that the indoor temperature is always controlled between the start value and the stop value. The calculation expression of the actual output power is:

[0057] P output =P max ·(1-ΔT relativ e·P adjust )

[0058] , where P outputis the actual output power of the air conditioner, used to maintain the indoor temperature, P max The maximum output power of the air conditioner, which indicates the energy output capacity of the air conditioner under full load;

[0059] The actual output power P of the air conditioner is adjusted dynamically output , control the indoor temperature between the start value and the stop value.

[0060] The multi-dimensional energy-saving analysis and dynamic management platform for buildings includes a temperature monitoring module, a temperature start-stop control module, a temperature data analysis module, a temperature change prediction module, a temperature change stage division module, a conventional start-stop control module, and a fuzzy logic power regulation module;

[0061] Temperature monitoring module, which monitors the indoor temperature in real time through the temperature monitoring system and transmits the collected temperature data to the control system;

[0062] The temperature start-stop control module compares the current indoor temperature with the preset start value and stop value of the air-conditioning system, and controls the start and stop of the air-conditioning system through the control system;

[0063] The temperature data analysis module collects real-time indoor temperature data and constructs it into an analysis data set, from which key features that can reflect the temperature change trend are extracted;

[0064] The temperature change prediction module conducts in-depth analysis of the extracted features under the monitoring window and inputs the processed data into the pre-trained machine learning model to identify and understand the laws of indoor temperature changes and predict future temperature change trends;

[0065] The temperature change stage division module predicts the indoor temperature change through a machine learning model and divides the indoor temperature change into a normal temperature change stage and a rapid temperature change stage;

[0066] Conventional start-stop control module: During the normal temperature change stage, the air conditioning system continues to monitor the current indoor temperature, compares it with the set start and stop values, and controls the start and stop of the air conditioning in a conventional manner;

[0067] The fuzzy logic power regulation module uses fuzzy logic to control the output power of the air conditioner according to the indoor temperature changes during the rapid temperature change stage, so that the current indoor temperature is always between the start value and the stop value, ensuring that the air conditioner always remains in operation, maintaining the indoor temperature while avoiding frequent start and stop of the air conditioner.

[0068] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0069] The present invention can accurately divide the temperature change stages through the collection and analysis of real-time temperature data combined with the prediction of the machine learning model, so as to adopt conventional start-stop control in the normal temperature change stage to reduce the system complexity and energy consumption; in the rapid temperature change stage, the air-conditioning output power is dynamically adjusted through fuzzy logic to ensure the continuous operation of the air-conditioning, maintain the room temperature stable and avoid the energy waste and equipment wear caused by the frequent start-up of the compressor. It not only improves the comfort of the indoor environment and the temperature control accuracy, but also significantly improves the energy efficiency of the system, reduces the operating cost, and extends the service life of the air-conditioning equipment, providing a high-efficiency and intelligent solution for building energy-saving management. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0071] Figure 1 This is a method flow chart of the multi-dimensional energy-saving analysis dynamic management method for buildings of the present invention.

[0072] Figure 2 It is a module schematic diagram of the multi-dimensional energy-saving analysis dynamic management platform for buildings of the present invention. DETAILED DESCRIPTION

[0073] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of the present disclosure will be more comprehensive and complete, and the concept of the example embodiments will be fully conveyed to those skilled in the art.

[0074] The present invention provides Figure 1 The multi-dimensional energy-saving analysis and dynamic management method of the building shown includes the following steps:

[0075] Monitor the indoor temperature in real time through the temperature monitoring system and transmit the collected temperature data to the control system;

[0076] The current indoor temperature is compared with the preset start value and stop value of the air conditioning system, and the start and stop of the air conditioning system are controlled by the control system;

[0077] The temperature monitoring system is a plurality of temperature sensors evenly arranged in the indoor area. The temperature data is collected through the plurality of temperature sensors, and the average value of the data collected by the plurality of temperature sensors is calculated to obtain the current indoor temperature.

[0078] By evenly arranging multiple temperature sensors in the indoor area and calculating the average value of the data collected by these sensors to obtain the current indoor temperature, the accuracy and reliability of temperature measurement can be significantly improved. Indoor temperature distribution is usually uneven. Affected by factors such as air flow, heat source distribution, wall insulation, window orientation, etc., a single temperature sensor may only reflect the temperature of a local area and cannot fully represent the actual temperature conditions of the entire room or area. Through the data fusion of multiple temperature sensors, these local differences can be eliminated and a more representative temperature value can be obtained. In addition, multiple temperature sensors can also improve the stability and fault tolerance of the system. Even if a temperature sensor fails or data deviation occurs, the data of other temperature sensors can still ensure the accuracy of overall temperature monitoring. This method can not only improve the control accuracy of the air-conditioning system and avoid instability caused by temperature fluctuations or temperature sensor errors, but also effectively reduce energy waste, optimize the air-conditioning adjustment strategy, and improve indoor comfort and energy efficiency.

[0079] Based on the real-time temperature data collected, the control system will compare and analyze the current indoor temperature with the preset start temperature value and stop temperature value of the air conditioning system. This process determines whether the air conditioning is started or stopped. The set start value and stop value are usually set according to the user's comfort needs and the design of the air conditioning system. For example, the start value is usually lower than the stop value to ensure that the air conditioning is turned on in advance before the temperature drops to the set value.

[0080] Collect and construct real-time indoor temperature data into an analytical dataset, from which key features that can reflect temperature change trends are extracted;

[0081] Under the monitoring window, the extracted features are deeply analyzed and the processed data is input into the pre-trained machine learning model to identify and understand the law of indoor temperature changes and predict future temperature change trends;

[0082] Extract key features that can reflect the trend of temperature changes, where the extracted key features include the speed at which the indoor real-time temperature deviates from the set value and the frequency and intensity of the real-time temperature fluctuation. After acquisition, in the monitoring window, conduct in-depth analysis on the speed at which the indoor real-time temperature deviates from the set value and the frequency and intensity of the real-time temperature fluctuation, and generate temperature deviation speed quantification values ​​and temperature fluctuation quantification values ​​respectively. The temperature deviation speed quantification value reflects the rapidity of temperature change; the temperature fluctuation quantification value reflects the volatility and amplitude of temperature change.

[0083] After obtaining the temperature deviation rate quantification value and temperature fluctuation quantification value generated by in-depth analysis of the extracted key features, the temperature deviation rate quantification value and the temperature fluctuation quantification value are input into a pre-trained machine learning model to generate a dynamic temperature change coefficient, and the future temperature change trend is predicted through the dynamic temperature change coefficient.

[0084] Pre-trained machine learning models are mainly used to analyze and predict indoor temperature trends. The model is usually trained based on historical temperature data, environmental factors, and known temperature control system behaviors. It learns the pattern, volatility, and response characteristics of temperature changes to identify complex temperature change patterns. During the training process, the model captures the relationship between key features such as the speed at which the temperature deviates from the set value and the frequency and intensity of temperature fluctuations and the temperature change trend. The training data can cover a variety of scenarios such as different indoor environments, external weather changes, and equipment working conditions, thereby helping the model to have a certain degree of generalization ability, so that even in the face of unknown situations in actual use, the model can make accurate predictions. Common machine learning algorithms include regression models (such as linear regression, support vector regression), ensemble learning methods (such as random forests, XGBoost), and deep learning methods (such as neural networks). These algorithms can efficiently extract features when processing large amounts of historical data and establish mathematical relationships between temperature changes and various factors.

[0085] During the application process, after the temperature deviation speed quantification value and the temperature fluctuation quantification value are input into the pre-trained machine learning model, the model will calculate a "dynamic temperature change coefficient" based on these input features. The dynamic temperature change coefficient quantifies the possible trend of future temperature changes. The dynamic temperature change coefficient reflects the direction, amplitude and time scale of indoor temperature changes, and can predict whether the future temperature will continue to be stable, fluctuate rapidly, or show a trend of gradually recovering or decreasing in the short term. In this way, the control system can obtain the predicted results of temperature changes in advance, thereby providing accurate information for temperature regulation. The dynamic temperature change coefficient provides a basis for the system to respond dynamically, ensuring that when there are large fluctuations in indoor temperature, the control system can respond in time, reduce the frequency and intensity of temperature fluctuations through adaptive adjustments, and thus maintain a comfortable indoor environment. With the predictive ability of the machine learning model, the system can not only provide more intelligent and efficient temperature control management under complex environmental changes, but also reduce energy consumption and avoid the burden of frequent start-stop on the air conditioning system.

[0086] The speed at which the real-time indoor temperature deviates from the set value is fast, which usually indicates that the indoor temperature is changing rapidly. The speed at which the temperature deviates from the set value (i.e., the temperature change rate) reflects the speed of temperature fluctuations. When the temperature changes rapidly, it means that environmental factors (such as changes in external weather, changes in air conditioning system load, or the influence of other heat sources) are rapidly affecting the indoor temperature. This rapid change may make it difficult for the temperature control system to adjust in time, which may cause the temperature to be unstable or exceed the comfort range. Generally, if the deviation rate of the indoor temperature exceeds the preset threshold, it means that the temperature change is beyond the normal range and immediate temperature control adjustment may be required. Therefore, the rapid temperature deviation rate is an important indicator of temperature changes, which can help the intelligent temperature control system identify and quickly respond to temperature changes and avoid fluctuations in environmental comfort.

[0087] In the monitoring window, the specific steps for in-depth analysis of the speed at which the indoor real-time temperature deviates from the set value to generate a quantitative value of the temperature deviation speed are as follows:

[0088] In the monitoring window, record the real-time indoor temperature at each time point and the target temperature set by the air conditioning system, and calculate the temperature deviation rate at each time point. The calculation expression is:

[0089]

[0090] , where T(t) is the real-time indoor temperature at time t, T s is the target temperature set for the air conditioner, R(t) is the temperature deviation rate at time t, expressed as a percentage;

[0091] By calculating the relative deviation of the real-time temperature from the target temperature, the temperature change at each moment is quantified in percentage form. Relative deviation is more versatile than absolute difference and can adapt to the range of changes in different target temperatures.

[0092] At each time point, the rate of change of the temperature deviation rate relative to time, that is, the deviation rate acceleration, is calculated. The calculation expression is:

[0093]

[0094] , where A(t) is the deviation rate acceleration at time t, indicating the speed of temperature deviation rate change, and Δt is the time interval between adjacent time points, in seconds;

[0095] By calculating the rate of change of the temperature deviation rate, the trend strength of the temperature change can be captured. If the deviation rate acceleration A(t) is large, it means that the speed of temperature deviation change is accelerating, indicating that the indoor temperature is in a state of rapid change.

[0096] The temperature deviation velocity quantification value is calculated based on the deviation rate acceleration to quantify the rapidity of temperature change. The formula is:

[0097]

[0098] , where TDSI xy is the quantized value of temperature deviation rate, and the α and β weighting factors are used to control the sensitivity to the deviation rate (i.e., the deviation rate acceleration A(t)) and the temperature deviation rate, e -β·|R(t)| is an exponential decay term. The higher the temperature deviation rate, the stronger the inhibition effect. t0 and t1 are the starting time point and the end time point of the monitoring window, respectively.

[0099] The setting of parameters α and β needs to be determined through experimental adjustment in combination with the actual application scenario to achieve reasonable sensitivity to the deviation rate and deviation ratio:

[0100] The influence of the deviation rate acceleration in the α control formula on the quantized value of the temperature deviation rate. A larger α value will enhance the system's response to rapid temperature changes, making the system more sensitive, and is suitable for scenarios that require rapid response, such as areas with drastic fluctuations in ambient temperature or high frequency of air conditioning startup. A smaller α value is suitable for environments with stable temperature changes, avoiding over-response and unnecessary adjustments. It is recommended to set the initial value in the range of 1-5 and gradually adjust it according to the actual operating status of the system.

[0101] β is used to adjust the sensitivity of the exponential decay term in the formula to the temperature deviation rate. A larger β value will make the exponential decay faster, thereby reducing the weight when the deviation rate is large, which is suitable for control systems that prioritize small-range temperature fluctuations. A smaller β value will amplify the impact of the deviation rate and is more suitable for environments with a large temperature range. It is recommended to set the initial value to the range of 0.1-1, observe the performance of the system under different temperature change conditions through experiments, and gradually optimize.

[0102] Ultimately, the adjustment of α and β needs to target the stability, sensitivity, and regulation accuracy of the system, and be optimized through experiments and data analysis according to the actual operating environment (such as external climate conditions and indoor load fluctuations) to ensure that the system can both respond quickly to rapid temperature changes and maintain precise regulation under stable conditions.

[0103] It can be seen from the temperature deviation speed quantification value that, under the monitoring window, the larger the performance value of the temperature deviation speed quantification value generated by in-depth analysis of the speed at which the indoor real-time temperature deviates from the set value, the faster the indoor temperature deviates from the set value, and the faster the indoor temperature changes; conversely, the smaller the performance value of the temperature deviation speed quantification value, the slower the temperature changes and the closer it is to a stable state. The temperature deviation speed quantification value combines the time rate of change of the temperature deviation rate (deviation rate acceleration) and the current temperature deviation degree, and can quantitatively evaluate the dynamic characteristics of temperature changes. When the temperature deviation speed quantification value is high, the system can determine that the temperature is changing rapidly, such as when the external environment fluctuates violently or when the indoor heat source changes significantly; and when the temperature deviation speed quantification value is low, it means that the temperature is basically in a stable state.

[0104] A high frequency and intensity of indoor real-time temperature fluctuations usually indicates that the indoor temperature is changing rapidly. A high frequency means that the temperature fluctuates frequently in a short period of time, while a high intensity reflects that the amplitude of each fluctuation is high. These characteristics are usually caused by factors such as changes in the external environment (such as sudden changes in external temperature or strong wind infiltration), indoor equipment operation (such as sudden start or stop of air conditioning and heating equipment), or drastic changes in personnel density and activities. When the frequency and intensity increase at the same time, it indicates that the indoor temperature has not remained stable, but is changing rapidly and significantly. Therefore, abnormal fluctuations in frequency and intensity are important signals for judging rapid temperature changes.

[0105] In the monitoring window, the specific steps for in-depth analysis of the frequency and intensity of indoor real-time temperature fluctuations to generate a quantitative value of temperature fluctuations are as follows:

[0106] The time series feature extraction of indoor real-time temperature fluctuation data is performed. Assume that the temperature data of N time points are collected in the monitoring window, which is recorded as T i ={T1, T2, ..., T N}, where T i is the temperature value at the i-th time point;

[0107] In order to analyze the frequency and intensity of temperature fluctuations, the amplitude of temperature changes at adjacent time points is first calculated to obtain the temperature difference series. The calculation expression of the temperature difference series is:

[0108] ΔT i =T i+1 -T i

[0109] , where ΔT i represents the i-th value of the temperature difference sequence, which represents the temperature variation amplitude of two adjacent sampling points, i=1, 2, 3, ..., N-1;

[0110] The temperature difference series can capture the dynamic characteristics of temperature changes over time. The goal of this step is to identify the core information of temperature fluctuations, eliminate the trend influence of absolute temperature, and only focus on the characteristics of fluctuation amplitude.

[0111] In order to quantify the intensity of fluctuations, a cumulative fluctuation energy function is defined, which combines the sum of squares of the difference sequence and the time weight to measure the overall severity of temperature fluctuations. The expression of the cumulative fluctuation energy function is:

[0112]

[0113] , where E T The accumulated fluctuation energy represents the intensity of temperature fluctuation within the monitoring window, w i is the time weight factor, which is used to highlight the contribution of fluctuations at different moments to the overall energy;

[0114] The weight factor can be set dynamically according to the following rules:

[0115]

[0116] , where is the sensitivity coefficient, which controls the effect of time on energy calculation. The larger the value, the more sensitive it is to the fluctuation of the center position of the monitoring window;

[0117] By accumulating the energy of fluctuations, we not only focus on the amplitude, but also reflect the time distribution characteristics of the fluctuations.

[0118] The temperature fluctuation quantification value is calculated by combining the accumulated fluctuation energy and the consistency of the temperature change direction to comprehensively reflect the amplitude of the fluctuation. The comprehensive calculation expression of the temperature fluctuation quantification value is:

[0119]

[0120] , where, TBI xy Indicates the quantitative value of temperature fluctuation, reflecting the volatility and amplitude of temperature change, sign(ΔT i ΔT i+1 ) indicates that the adjacent temperature difference (ΔT i ΔT i+1 ) is consistent to calculate the directional consistency of temperature change. If the sign consistency is high, it means that the fluctuation is relatively smooth; if the consistency is low, it means that the fluctuation is more complex. δ is the directional consistency weight coefficient, which is used to balance the contribution of fluctuation intensity and directional consistency.

[0121] It can be seen from the temperature fluctuation quantification value that, under the monitoring window, the larger the performance value of the temperature fluctuation quantification value generated by in-depth analysis of the frequency and intensity of indoor real-time temperature fluctuations, the faster the indoor temperature changes; conversely, the smaller the performance value, the slower the indoor temperature changes. The temperature fluctuation quantification value is an indicator generated by comprehensive analysis of the frequency and intensity of temperature fluctuations, and its value directly reflects the severity of temperature changes. A larger temperature fluctuation quantification value means that within the monitoring window, the temperature changes more frequently and the fluctuation amplitude is larger, while a smaller index value indicates that the temperature changes smoothly, with lower fluctuation amplitude and frequency. Therefore, the temperature fluctuation quantification value is a key indicator for quantifying the speed of indoor temperature changes, and provides basic support for further predicting temperature change trends.

[0122] The machine learning model is not specifically limited here, and can achieve the quantification of temperature deviation speed TDSI xy and temperature fluctuation quantification value TBI xy Perform comprehensive analysis to generate dynamic temperature change coefficient Temper dyna In order to realize the technical solution of the present invention, the present invention provides a specific implementation method; Dynamic temperature change coefficient Temper dyna The generated expression is:

[0123]

[0124] , where k1 and k2 are the temperature deviation speed quantization values ​​TDSI respectively. xy and temperature fluctuation quantification value TBI xy The preset proportional coefficient, and k1 and k2 are both greater than 0.

[0125] It can be seen from the dynamic temperature change coefficient that, under the monitoring window, the greater the performance value of the temperature deviation speed quantization value generated by in-depth analysis of the speed at which the indoor real-time temperature deviates from the set value, the greater the performance value of the temperature fluctuation quantization value generated by in-depth analysis of the frequency and intensity of the indoor real-time temperature fluctuations. That is, the greater the performance value of the dynamic temperature change coefficient generated when predicting the indoor temperature change trend through the pre-trained machine learning model under the monitoring window, the faster the dynamic change of the indoor temperature, and vice versa.

[0126] The indoor temperature change is predicted through the machine learning model, and the indoor temperature change is divided into the normal temperature change stage and the rapid temperature change stage;

[0127] The dynamic temperature change coefficient generated by the pre-trained machine learning model when predicting the indoor temperature change trend under the monitoring window is compared and analyzed with the pre-set dynamic temperature change coefficient reference threshold, and the indoor temperature change is divided. The specific division steps are as follows:

[0128] If the dynamic temperature change coefficient generated when predicting the indoor temperature change trend through the pre-trained machine learning model under the monitoring window is greater than the dynamic temperature change coefficient reference threshold, the indoor temperature change is divided into a rapid temperature change stage;

[0129] If the dynamic temperature change coefficient generated when predicting the indoor temperature change trend through the pre-trained machine learning model under the monitoring window is less than or equal to the dynamic temperature change coefficient reference threshold, the indoor temperature change is classified as the normal temperature change stage.

[0130] The purpose of dividing the temperature change window period is to enable the air conditioner to respond more accurately to temperature changes under different environmental conditions.

[0131] During the normal temperature change phase, the air conditioning system continues to monitor the current indoor temperature, compares it with the set start and stop values, and controls the start and stop of the air conditioning in the usual way;

[0132] During the normal temperature change stage, the air conditioning system continuously monitors the current indoor temperature and compares it with the set start and stop values. The air conditioning system starts and stops in the conventional way, which can effectively avoid unnecessary frequent adjustments while maintaining the stability of the indoor temperature. The advantage of this method is that it can not only meet the user's temperature comfort needs, but also reduce the system's operating load and extend the service life of the air conditioning equipment. In addition, the conventional control method can avoid complex algorithm calculations and unnecessary over-responses at this stage, thereby reducing energy consumption and achieving a balance between economy and comfort. Through precise and simple control logic, the air conditioning system can maintain efficient operation during the normal temperature change stage, ensuring the best combination of energy consumption and environmental comfort.

[0133] In the stage of rapid temperature change, fuzzy logic is used to control the output power of the air conditioner according to the indoor temperature change, so that the current indoor temperature is always between the start value and the stop value, ensuring that the air conditioner always remains in operation, maintaining the indoor temperature while avoiding frequent start and stop of the air conditioner;

[0134] In the stage of rapid temperature change, fuzzy logic is used to control the output power of the air conditioner according to the indoor temperature change, so that the current indoor temperature is always between the start value and the stop value. The specific steps are as follows:

[0135] In the rapid temperature change stage, the dynamic temperature change coefficient is first compared with the dynamic temperature change coefficient reference threshold to determine the severity of the current temperature change, and as one of the input variables of the fuzzy logic, the relative position between the current indoor temperature and the start value and stop value of the air conditioner is calculated to generate the temperature deviation factor. The expression for generating the temperature deviation factor is:

[0136]

[0137] , where ΔT relative is the temperature deviation factor, which quantifies the degree of deviation of the current temperature within the target range. The temperature deviation factor ranges from [0, 1] and indicates the degree of deviation of the current temperature relative to the start value and the stop value. current is the current indoor temperature, T start is the starting value when the air conditioner is started, T stop It is the stop value when the air conditioner stops;

[0138] Calculate the normalized value of the dynamic temperature change amplitude. The calculation expression is:

[0139]

[0140] , among which, Temper dyna Indicates the dynamic temperature change coefficient, DTC_ref indicates the dynamic temperature change coefficient reference threshold, ΔT dynamic is a normalized value of dynamic temperature change, quantifying the ratio of the dynamic temperature change coefficient to a reference threshold value of the dynamic temperature change coefficient;

[0141] Temperature deviation factor ΔT relative and the normalized value of dynamic temperature change ΔT dynamic It will serve as the core input of the fuzzy logic system to analyze the drastic change of the current temperature and its position in the target range.

[0142] Temperature deviation factor ΔT based on input relative and the normalized value of dynamic temperature change ΔT dynamic , the air conditioner output power adjustment coefficient is generated by fuzzy logic rules. The fuzzy logic rules consider two aspects:

[0143] When the temperature is close to the start value or stop value (temperature deviation factor ΔT relative close to 0 or 1) and the temperature changes dramatically (dynamic temperature change normalized value ΔT dynamic >1), increase the output power to quickly stabilize the temperature;

[0144] When the temperature is far away from the start value and the stop value (temperature deviation factor ΔT relative In the middle range) and the temperature changes more smoothly (dynamic temperature change normalized value ΔT dynamic ≤1), reduce the power to avoid excessive response;

[0145] The calculation expression of fuzzy logic output is:

[0146] P adjust =min(1,max(0,k pΔT relative +k q ΔT dynamic ))

[0147] , where k p and k q are weight coefficients, which respectively control the influence of relative temperature deviation and dynamic change on output power adjustment, P adjust The output power adjustment coefficient of the air conditioner is used to dynamically adjust the output power of the air conditioner. The min and max values ​​are used to limit the value range of the output power adjustment coefficient to ensure safety.

[0148] The actual output power of the air conditioner is calculated based on the generated adjustment coefficient. At this stage, it is necessary to ensure that the indoor temperature is always controlled between the start value and the stop value. Therefore, the actual output power is calculated by the following formula:

[0149] P output =P max ·(1-ΔT relative ·P adjust )

[0150] , where P output is the actual output power of the air conditioner, used to maintain the indoor temperature, P max The maximum output power of the air conditioner, which indicates the energy output capacity of the air conditioner under full load;

[0151] Finally, the actual output power P of the air conditioner is adjusted dynamically. output , the air conditioner can maintain continuous operation during the rapid temperature change stage, stably control the indoor temperature between the start value and the stop value, and avoid energy waste and system pressure caused by frequent start and stop.

[0152] In the stage of rapid temperature changes, fuzzy logic is used to control the output power of the air conditioner according to the indoor temperature changes. Its specific role is to ensure that the indoor temperature is always between the set start value and stop value by dynamically adjusting the output power, thereby avoiding unnecessary energy consumption and equipment wear caused by frequent start and stop of the air conditioner. This method can keep the air conditioner running continuously when the temperature changes drastically, and achieve stable temperature control through refined power adjustment. On the one hand, it can quickly respond to environmental changes and prevent the indoor temperature from deviating from the comfortable range; on the other hand, the frequent start and stop of the compressor is avoided through the gradual adjustment of power, thereby extending the service life of the air conditioner, while improving the system energy efficiency and ensuring the comfort and economy of the user.

[0153] The present invention can accurately divide the temperature change stages through the collection and analysis of real-time temperature data combined with the prediction of the machine learning model, so as to adopt conventional start-stop control in the normal temperature change stage to reduce the system complexity and energy consumption; in the rapid temperature change stage, the air-conditioning output power is dynamically adjusted through fuzzy logic to ensure the continuous operation of the air-conditioning, maintain the room temperature stable and avoid the energy waste and equipment wear caused by the frequent start-up of the compressor. It not only improves the comfort of the indoor environment and the temperature control accuracy, but also significantly improves the energy efficiency of the system, reduces the operating cost, and extends the service life of the air-conditioning equipment, providing a high-efficiency and intelligent solution for building energy-saving management.

[0154] The present invention provides Figure 2 The multi-dimensional energy-saving analysis dynamic management platform of the building shown includes a temperature monitoring module, a temperature start-stop control module, a temperature data analysis module, a temperature change prediction module, a temperature change stage division module, a conventional start-stop control module and a fuzzy logic power regulation module;

[0155] Temperature monitoring module, which monitors the indoor temperature in real time through the temperature monitoring system and transmits the collected temperature data to the control system;

[0156] The temperature start-stop control module compares the current indoor temperature with the preset start value and stop value of the air-conditioning system, and controls the start and stop of the air-conditioning system through the control system;

[0157] The temperature data analysis module collects real-time indoor temperature data and constructs it into an analysis data set, from which key features that can reflect the temperature change trend are extracted;

[0158] The temperature change prediction module conducts in-depth analysis of the extracted features under the monitoring window and inputs the processed data into the pre-trained machine learning model to identify and understand the laws of indoor temperature changes and predict future temperature change trends;

[0159] The temperature change stage division module predicts the indoor temperature change through a machine learning model and divides the indoor temperature change into a normal temperature change stage and a rapid temperature change stage;

[0160] Conventional start-stop control module: During the normal temperature change stage, the air conditioning system continues to monitor the current indoor temperature, compares it with the set start and stop values, and controls the start and stop of the air conditioning in a conventional manner;

[0161] The fuzzy logic power regulation module uses fuzzy logic to control the output power of the air conditioner according to the indoor temperature changes during the rapid temperature change stage, so that the current indoor temperature is always between the start value and the stop value, ensuring that the air conditioner remains in operation, maintaining the indoor temperature while avoiding frequent start and stop of the air conditioner;

[0162] The multi-dimensional energy-saving analysis dynamic management method for a building provided in an embodiment of the present invention is implemented through the multi-dimensional energy-saving analysis dynamic management platform for the above-mentioned building. The specific methods and processes of the multi-dimensional energy-saving analysis dynamic management platform for the building are detailed in the embodiment of the multi-dimensional energy-saving analysis dynamic management method for the above-mentioned building, which will not be repeated here.

[0163] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0164] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0165] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A multi-dimensional energy-saving analysis and dynamic management method for buildings, characterized in that: The following steps are involved: Monitor the indoor temperature in real time through the temperature monitoring system and transmit the collected temperature data to the control system; The current indoor temperature is compared with the preset start value and stop value of the air conditioning system, and the start and stop of the air conditioning system are controlled by the control system; Collect and construct real-time indoor temperature data into an analytical dataset, from which key features that can reflect temperature change trends are extracted; Under the monitoring window, the extracted features are deeply analyzed and the processed data is input into the pre-trained machine learning model to identify and understand the law of indoor temperature changes and predict future temperature change trends; The indoor temperature change is predicted through the machine learning model, and the indoor temperature change is divided into the normal temperature change stage and the rapid temperature change stage; During the normal temperature change phase, the air conditioning system continues to monitor the current indoor temperature, compares it with the set start and stop values, and controls the start and stop of the air conditioning in the usual way; During the rapid temperature change stage, fuzzy logic is used to control the output power of the air conditioner according to the indoor temperature changes, so that the current indoor temperature is always between the start value and the stop value, ensuring that the air conditioner always remains in operation, maintaining the indoor temperature while avoiding frequent start and stop of the air conditioner.

2. The multi-dimensional energy-saving analysis and dynamic management method for buildings according to claim 1 is characterized in that: The temperature monitoring system is a system that evenly distributes multiple temperature sensors in the indoor area. The system collects temperature data through multiple temperature sensors and calculates the average value of the data collected by the multiple temperature sensors to obtain the current indoor temperature.

3. The multi-dimensional energy-saving analysis and dynamic management method for buildings according to claim 1 is characterized in that: Extract key features that can reflect the temperature change trend, where the extracted key features include the speed at which the indoor real-time temperature deviates from the set value and the frequency and intensity of the real-time temperature fluctuation. After acquisition, in the monitoring window, conduct in-depth analysis on the speed at which the indoor real-time temperature deviates from the set value and the frequency and intensity of the real-time temperature fluctuation, and generate a temperature deviation speed quantification value and a temperature fluctuation quantification value respectively. The temperature deviation speed quantification value reflects the rapidity of the temperature change; the temperature fluctuation quantification value reflects the volatility and amplitude of the temperature change; After obtaining the temperature deviation rate quantification value and temperature fluctuation quantification value generated by in-depth analysis of the extracted key features, the temperature deviation rate quantification value and the temperature fluctuation quantification value are input into a pre-trained machine learning model to generate a dynamic temperature change coefficient, and the future temperature change trend is predicted through the dynamic temperature change coefficient.

4. The multi-dimensional energy-saving analysis and dynamic management method for buildings according to claim 3 is characterized in that: The dynamic temperature change coefficient generated by the pre-trained machine learning model when predicting the indoor temperature change trend under the monitoring window is compared and analyzed with the pre-set dynamic temperature change coefficient reference threshold, and the indoor temperature change is divided. The specific division steps are as follows: If the dynamic temperature change coefficient generated when predicting the indoor temperature change trend through the pre-trained machine learning model under the monitoring window is greater than the dynamic temperature change coefficient reference threshold, the indoor temperature change is divided into a rapid temperature change stage; If the dynamic temperature change coefficient generated when predicting the indoor temperature change trend through the pre-trained machine learning model under the monitoring window is less than or equal to the dynamic temperature change coefficient reference threshold, the indoor temperature change is classified as the normal temperature change stage.

5. The multi-dimensional energy-saving analysis and dynamic management method for buildings according to claim 3 is characterized in that: In the monitoring window, the specific steps for in-depth analysis of the speed at which the indoor real-time temperature deviates from the set value to generate a quantitative value of the temperature deviation speed are as follows: In the monitoring window, record the real-time indoor temperature at each time point and the target temperature set by the air conditioning system, and calculate the temperature deviation rate at each time point. The calculation expression is: , where T(t) is the real-time indoor temperature at time t, T s is the target temperature set for the air conditioner, R(t) is the temperature deviation rate at time t, expressed as a percentage; At each time point, the rate of change of the temperature deviation rate relative to time, that is, the deviation rate acceleration, is calculated. The calculation expression is: , where A(t) is the deviation rate acceleration at time t, indicating the speed of change of temperature deviation rate, and Δt is the time interval between adjacent time points; The temperature deviation velocity quantification value is calculated based on the deviation rate acceleration to quantify the rapidity of temperature change. The formula is: , where TDSI xy is the quantitative value of the temperature deviation rate, and the α and β weighting factors are used to control the sensitivity to the deviation rate and temperature deviation rate, e -β·|R(t)| is an exponential decay term, t0 and t1 are the starting time point and the ending time point of the monitoring window respectively.

6. The multi-dimensional energy-saving analysis and dynamic management method for buildings according to claim 3 is characterized in that: In the monitoring window, the specific steps for in-depth analysis of the frequency and intensity of indoor real-time temperature fluctuations to generate a quantitative value of temperature fluctuations are as follows: The time series feature extraction of indoor real-time temperature fluctuation data is performed. Assume that the temperature data of N time points are collected in the monitoring window, which is recorded as T i ={T1, T2, ..., T N }, where T i is the temperature value at the i-th time point; Calculate the amplitude of temperature change at adjacent time points to obtain the temperature difference sequence. The calculation expression of the temperature difference sequence is: ΔT i =T i+1 -T i , where ΔT i represents the i-th value of the temperature difference sequence, which represents the temperature variation amplitude of two adjacent sampling points, i=1, 2, 3, ..., N-1; A cumulative fluctuation energy function is defined to quantify the fluctuation intensity. It is combined with the sum of squares of the difference sequence and the time weight to measure the overall severity of the temperature fluctuation. The expression of the cumulative fluctuation energy function is: , where E T The accumulated fluctuation energy represents the intensity of temperature fluctuation within the monitoring window, w i is the time weight factor, which is used to highlight the contribution of fluctuations at different moments to the overall energy; The temperature fluctuation quantification value is calculated by combining the accumulated fluctuation energy and the consistency of the temperature change direction to comprehensively reflect the amplitude of the fluctuation. The comprehensive calculation expression of the temperature fluctuation quantification value is: , where, TBI xy Indicates the quantitative value of temperature fluctuation, reflecting the volatility and amplitude of temperature change, sign(ΔT i ΔT i+1 ) indicates that the adjacent temperature difference (ΔT i ΔT i+1 ) is consistent to calculate the directional consistency of temperature change. δ is the directional consistency weight coefficient, which is used to balance the contribution of fluctuation intensity and directional consistency.

7. The multi-dimensional energy-saving analysis and dynamic management method for buildings according to claim 4 is characterized in that: In the stage of rapid temperature change, fuzzy logic is used to control the output power of the air conditioner according to the indoor temperature change, so that the current indoor temperature is always between the start value and the stop value. The specific steps are as follows: In the rapid temperature change stage, the dynamic temperature change coefficient is first compared with the dynamic temperature change coefficient reference threshold to determine the severity of the current temperature change, and as one of the input variables of the fuzzy logic, the relative position between the current indoor temperature and the start value and stop value of the air conditioner is calculated to generate the temperature deviation factor. The expression for generating the temperature deviation factor is: , where ΔT relative is the temperature deviation factor, which quantifies the degree of deviation of the current temperature within the target range. The temperature deviation factor ranges from [0, 1] and indicates the degree of deviation of the current temperature relative to the start value and the stop value. current is the current indoor temperature, T start is the starting value when the air conditioner is started, T stop It is the stop value when the air conditioner stops; Calculate the normalized value of the dynamic temperature change amplitude. The calculation expression is: , among which, Temper dyna Indicates the dynamic temperature change coefficient, DTC_ref indicates the dynamic temperature change coefficient reference threshold, ΔT dynamic is a normalized value of dynamic temperature change, quantifying the ratio of the dynamic temperature change coefficient to a reference threshold value of the dynamic temperature change coefficient; The temperature deviation factor ΔT relative and the normalized value of dynamic temperature change ΔT dynamic It will serve as the core input of the fuzzy logic system to analyze the drastic change of the current temperature and its position in the target range.

8. The multi-dimensional energy-saving analysis and dynamic management method for buildings according to claim 7 is characterized in that: Temperature deviation factor ΔT based on input relative and the normalized value of dynamic temperature change ΔT dynamic , the air conditioner output power adjustment coefficient is generated by fuzzy logic rules. The fuzzy logic rules consider two aspects: When the temperature is close to the start value or stop value and the temperature changes dramatically, increase the output power to quickly stabilize the temperature; When the temperature is far away from the start value and the stop value and the temperature changes smoothly, reduce the power to avoid excessive response; The calculation expression of fuzzy logic output is: P adjust =min((1,max(0,k p ·ΔT relatiue +k q ·ΔT dynamic )) , where k p and k q are weight coefficients, which respectively control the influence of relative temperature deviation and dynamic change on output power adjustment, P adjust is the air conditioner output power adjustment coefficient, which is used to dynamically adjust the output power of the air conditioner. Min and max are used to limit the value range of the output power adjustment coefficient.

9. The multi-dimensional energy-saving analysis and dynamic management method for buildings according to claim 8 is characterized in that: The actual output power of the air conditioner is calculated based on the generated adjustment coefficient. At this stage, it is necessary to ensure that the indoor temperature is always controlled between the start value and the stop value. The calculation expression of the actual output power is: P output =P max ·(1-ΔT relative ·P adjust ) , where P output is the actual output power of the air conditioner, used to maintain the indoor temperature, P max The maximum output power of the air conditioner, indicating the energy output capacity of the air conditioner under full load; The actual output power P of the air conditioner is adjusted dynamically output , control the indoor temperature between the start value and the stop value.

10. A multi-dimensional energy-saving analysis dynamic management platform for buildings, used to implement the multi-dimensional energy-saving analysis dynamic management method for buildings described in any one of claims 1 to 9, characterized in that: It includes a temperature monitoring module, a temperature start-stop control module, a temperature data analysis module, a temperature change prediction module, a temperature change stage division module, a conventional start-stop control module and a fuzzy logic power regulation module; Temperature monitoring module, which monitors the indoor temperature in real time through the temperature monitoring system and transmits the collected temperature data to the control system; The temperature start-stop control module compares the current indoor temperature with the preset start value and stop value of the air-conditioning system, and controls the start and stop of the air-conditioning system through the control system; The temperature data analysis module collects real-time indoor temperature data and constructs it into an analysis data set, from which key features that can reflect the temperature change trend are extracted; The temperature change prediction module conducts in-depth analysis of the extracted features under the monitoring window and inputs the processed data into the pre-trained machine learning model to identify and understand the laws of indoor temperature changes and predict future temperature change trends; The temperature change stage division module predicts the indoor temperature change through a machine learning model and divides the indoor temperature change into a normal temperature change stage and a rapid temperature change stage; Conventional start-stop control module: During the normal temperature change stage, the air conditioning system continues to monitor the current indoor temperature, compares it with the set start and stop values, and controls the start and stop of the air conditioning in a conventional manner; The fuzzy logic power regulation module uses fuzzy logic to control the output power of the air conditioner according to the indoor temperature changes during the rapid temperature change stage, so that the current indoor temperature is always between the start value and the stop value, ensuring that the air conditioner always remains in operation, maintaining the indoor temperature while avoiding frequent start and stop of the air conditioner.

Images