Energy-saving and emission-reducing building management system and method

Abstract

The invention relates to the technical field of building management, and particularly discloses an energy-saving and emission-reducing building management system and method, and the system comprises the steps: carrying out the time sequence analysis of indoor and outdoor temperature data and humidity data through employing an artificial intelligence technology based on deep learning, and extracting the periodic change characteristics of an indoor and outdoor temperature difference value and a humidity difference value; and through further kernel feature extraction, a core feature mode of time sequence change of indoor and outdoor temperature difference and humidity difference is mined, so that intelligent recommendation and optimal control on the temperature set value of the air conditioner in the building are realized on the basis. Thus, the working state of the air conditioner is automatically adjusted by monitoring indoor and outdoor temperature and humidity changes in real time, the comfort degree of the indoor environment can be effectively maintained, energy consumption is reduced, and therefore the purposes of energy conservation and emission reduction are achieved.

Description

Technical Field

[0001] The present application relates to the technical field of building management, and more specifically, to a building management system and method for energy conservation and emission reduction. Background Art

[0002] With the increasingly severe global climate change and the continuous growth of energy consumption, energy conservation and emission reduction in the construction field has become one of the focuses of international attention. According to relevant statistics, the construction industry accounts for a considerable proportion of the total energy consumption, and the energy consumption of air-conditioning systems accounts for the main part of the total energy consumption of buildings. Therefore, improving the energy efficiency of building air-conditioning systems and reducing unnecessary energy waste are of great significance to reducing building operating costs and promoting environmental protection.

[0003] Traditional building air conditioning control systems mostly rely on preset schedules or simple temperature thresholds to adjust the working status of the air conditioner. This control method often lacks flexibility and accuracy, and cannot effectively respond to changes in indoor and outdoor environmental conditions, which can easily lead to excessive use of energy or a decrease in indoor comfort.

[0004] Therefore, an optimized building management system for energy conservation and emission reduction is expected. Summary of the invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a building management system and method for energy conservation and emission reduction, which uses artificial intelligence technology based on deep learning to perform time series analysis on indoor and outdoor temperature data and humidity data, respectively extracts the periodic change characteristics of the indoor and outdoor temperature difference and humidity difference, and further refines the kernel features to dig out the core feature patterns of the time series changes of the indoor and outdoor temperature difference and humidity difference, so as to realize intelligent recommendation and optimization control of the air-conditioning temperature setting value in the building on this basis. In this way, by monitoring the indoor and outdoor temperature and humidity changes in real time to automatically adjust the working state of the air conditioner, the comfort of the indoor environment can be effectively maintained and energy consumption can be reduced, thereby achieving the purpose of energy conservation and emission reduction.

[0006] Accordingly, according to one aspect of the present application, there is provided a building management system for energy saving and emission reduction, which includes: A temperature data acquisition module, used for receiving a time queue of indoor temperature values ​​and a time queue of outdoor temperature values ​​collected by a temperature sensor; A humidity data acquisition module, used for receiving a time queue of indoor humidity values ​​and a time queue of outdoor humidity values ​​collected by a humidity sensor; An environmental difference periodicity analysis module, configured to perform periodicity analysis of indoor and outdoor environmental differences based on the time queue of the indoor temperature value and the time queue of the outdoor temperature value, the time queue of the indoor humidity value and the time queue of the outdoor humidity value to obtain a plurality of indoor and outdoor temperature difference subsequences and a plurality of indoor and outdoor humidity difference subsequences; A kernel feature extraction module, used for performing kernel feature extraction based on multidimensional energy potential analysis on the multiple indoor and outdoor temperature difference subsequences and the multiple indoor and outdoor humidity difference subsequences to obtain indoor and outdoor temperature difference time series kernel feature vectors and indoor and outdoor humidity difference time series kernel feature vectors; The air conditioning temperature optimization management module is used to generate a management optimization instruction for the air conditioning temperature setting value based on the fusion characteristics of the indoor and outdoor temperature difference time series kernel feature vector and the indoor and outdoor humidity difference time series kernel feature vector.

[0007] Compared with the prior art, the building management system and method for energy conservation and emission reduction provided by the present application uses artificial intelligence technology based on deep learning to perform time series analysis on indoor and outdoor temperature data and humidity data, respectively extracts the periodic change characteristics of the indoor and outdoor temperature difference and humidity difference, and further refines the kernel features to dig out the core feature patterns of the time series changes of the indoor and outdoor temperature difference and humidity difference, so as to realize intelligent recommendation and optimization control of the air conditioning temperature setting value in the building on this basis. In this way, by monitoring the indoor and outdoor temperature and humidity changes in real time to automatically adjust the working state of the air conditioner, the comfort of the indoor environment can be effectively maintained and energy consumption can be reduced, thereby achieving the purpose of energy conservation and emission reduction. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0009] Figure 1 A block diagram of a building management system for energy conservation and emission reduction according to an embodiment of the present application.

[0010] Figure 2 Schematic diagram of data flow of a building management system for energy conservation and emission reduction according to an embodiment of the present application.

[0011] Figure 3 It is a block diagram of an environmental difference periodic analysis module in a building management system for energy conservation and emission reduction according to an embodiment of the present application.

[0012] Figure 4It is a block diagram of a kernel feature extraction module in a building management system for energy conservation and emission reduction according to an embodiment of the present application.

[0013] Figure 5 It is a block diagram of an air conditioning temperature optimization management module in a building management system for energy conservation and emission reduction according to an embodiment of the present application. DETAILED DESCRIPTION

[0014] Below, the embodiments of the present application will be described in more detail in conjunction with the accompanying drawings, and the above and other purposes, features and advantages of the present application will become more apparent. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the example embodiments described herein.

[0015] In response to the technical problems described in the above background technology, the present application proposes a building management system for energy saving and emission reduction. Figure 1 A block diagram of a building management system for energy conservation and emission reduction according to an embodiment of the present application. Figure 2 Schematic diagram of data flow of building management system for energy saving and emission reduction according to an embodiment of the present application. Figure 1 and Figure 2 As shown, the building management system 100 for energy conservation and emission reduction according to the embodiment of the present application includes: a temperature data acquisition module 110, which is used to receive a time queue of indoor temperature values ​​and a time queue of outdoor temperature values ​​collected by a temperature sensor; a humidity data acquisition module 120, which is used to receive a time queue of indoor humidity values ​​and a time queue of outdoor humidity values ​​collected by a humidity sensor; an environmental difference periodicity analysis module 130, which is used to perform indoor and outdoor environmental difference periodicity analysis based on the time queue of the indoor temperature value and the time queue of the outdoor temperature value, the time queue of the indoor humidity value and the time queue of the outdoor humidity value to obtain multiple indoor and outdoor temperature difference subsequences and multiple indoor and outdoor humidity difference subsequences; a kernel feature extraction module 140, which is used to perform kernel feature extraction based on multidimensional energy potential analysis on the multiple indoor and outdoor temperature difference subsequences and the multiple indoor and outdoor humidity difference subsequences to obtain indoor and outdoor temperature difference time series kernel feature vectors and indoor and outdoor humidity difference time series kernel feature vectors; an air conditioning temperature optimization management module 150, which is used to generate a management optimization instruction for the air conditioning temperature setting value based on the fusion features of the indoor and outdoor temperature difference time series kernel feature vectors and the indoor and outdoor humidity difference time series kernel feature vectors.

[0016] In the above-mentioned building management system 100 for energy conservation and emission reduction, the temperature data acquisition module 110 is used to receive the time queue of indoor temperature values ​​and the time queue of outdoor temperature values ​​collected by the temperature sensor. It should be understood that the indoor temperature is a key factor affecting human comfort, while the outdoor temperature reflects the external environmental heat load that the air-conditioning system needs to overcome. When the temperature difference between indoor and outdoor is too large, the air conditioner needs to consume more energy to maintain the indoor temperature. For example, in hot summer weather, if the outdoor temperature is very high and the indoor temperature is set too low, the air conditioner needs to run for a long time to lower the indoor temperature, which will lead to increased energy consumption. Therefore, by adjusting the air conditioner temperature setting in combination with the indoor temperature and the outdoor temperature, the air conditioner can reduce energy consumption while maintaining comfort, thereby achieving the purpose of energy conservation and emission reduction.

[0017] In this technical solution, receiving the time queue of indoor temperature values ​​and the time queue of outdoor temperature values ​​collected by the temperature sensor is a key data collection step, which not only involves the deployment of temperature sensors and the selection of data collection frequency, but also includes multiple links such as data transmission, storage, verification, cleaning, synchronization and backup. The following is a detailed description of this process: First of all, the deployment of temperature sensors is the basis of data collection. Inside the building, indoor temperature sensors need to be deployed in multiple areas indoors, such as walls, ceilings or floors. The comprehensive representation of indoor temperature is obtained by calculating the average value of temperature data collected by each sensor to ensure comprehensive coverage of the entire building space and the accuracy and representativeness of temperature data. In practical applications, the specific location of the temperature sensor should ensure that the sensor can accurately measure the temperature of each area while avoiding interference from other factors (such as heat sources, cold sources, etc.). Similarly, outside the building, it is also necessary to select a suitable location to deploy outdoor temperature sensors. Outdoor sensors should be placed in a location that can represent the ambient temperature and avoid being directly affected by factors such as direct sunlight and wind speed changes. For example, you can choose a shaded place or a place with fewer obstructions to ensure that the sensor can truly reflect the temperature changes of the outdoor environment.

[0018] Data collection frequency: The choice of data collection frequency needs to balance the real-time nature of the data and the storage cost. Typically, the collection frequency of indoor and outdoor temperature sensors can be set to once every minute or every five minutes. A higher collection frequency can provide more detailed temperature change data, which helps to more accurately analyze the differences between indoor and outdoor environments, but it also increases the burden of data storage and processing. Although a lower collection frequency reduces storage requirements, it may not capture temperature fluctuations in a short period of time. Therefore, the choice of collection frequency should be adjusted according to the needs of specific application scenarios.

[0019] Data transmission: The data collected by the temperature sensor needs to be transmitted to the data processor through reliable communication technology. Common data transmission methods include wireless transmission and wired transmission. Specifically, wireless communication technologies such as Wi-Fi, Zigbee, LoRa, etc. are widely used in data transmission of IoT devices. These technologies have the advantages of long transmission distance, low power consumption, and easy deployment. For example, Zigbee is suitable for short-distance and low-power scenarios, while LoRa is suitable for long-distance and low-bandwidth scenarios. Choosing a suitable wireless communication technology can ensure stable data transmission and reduce the risk of data loss. In some special scenarios, wired methods (such as RS485, Ethernet, etc.) can also be used to transmit data to the data processor. The advantages of wired transmission are fast transmission speed and high stability, but the wiring cost is high and it is not easy to maintain later. Therefore, wired transmission is usually used in occasions with high requirements for data transmission.

[0020] After the data is transferred to the data processor, effective storage management measures need to be implemented. Data storage can be divided into two levels: local storage and cloud storage. Specifically, in terms of local storage: the temperature sensor has the ability to store data for a certain period of time locally, aiming to prevent data loss caused by network failures. Local storage usually relies on built-in flash memory modules or external memory cards. Although its storage capacity is limited, it can play the role of temporary data cache to ensure data integrity. As for cloud storage, the data processor will save the received data to the cloud server, laying the foundation for subsequent data processing and analysis. Cloud storage shows unlimited expansion potential and high reliability, and can easily perform data backup and recovery operations. In addition, the cloud platform also provides a variety of data management and analysis tools, providing users with convenient conditions for data visualization and statistical analysis.

[0021] Data verification and cleaning are important links to ensure data quality. The received data needs to be strictly verified and cleaned to remove outliers and erroneous data to ensure the accuracy of subsequent analysis. Specifically, data verification refers to the verification of the received data to ensure the integrity and accuracy of the data. Specific methods include checking whether the timestamp of the data is continuous and whether the temperature value is within a reasonable range. For example, a reasonable temperature threshold can be set, and when the temperature value exceeds the threshold, it is marked as abnormal data. In addition, by comparing the data of multiple sensors, data inconsistencies can be found and corrected. Data cleaning refers to the removal of abnormal data, such as abnormally high or low temperature values ​​caused by sensor failure. Data cleaning can use statistical methods (such as mean, standard deviation, etc.) or machine learning algorithms for anomaly detection and data cleaning. For example, by calculating the standard deviation of temperature values ​​over a period of time, data points that deviate far from the standard deviation can be marked as outliers and deleted from the data set. In addition, sliding window technology can be used to smooth the data to reduce the impact of temperature fluctuations in a short period of time on the analysis results.

[0022] Data synchronization and backup are key measures to ensure data security and availability. By regularly synchronizing and backing up data, data loss or corruption can be prevented and the normal operation of the system can be ensured. Specifically, data synchronization can ensure that the data of all sensors can be synchronized to the data processor in a timely manner to avoid data delay or loss. Data synchronization can be achieved through scheduled tasks or event triggers. For example, data synchronization can be set to be synchronized every ten minutes, or immediately when sensor data changes. Data backup refers to the regular backup of data stored in the cloud to prevent data loss or corruption. Data backup can use the automatic backup function provided by the cloud platform, or it can be backed up manually. Backup data should be stored in different physical locations to prevent data loss due to a single point of failure. In addition, a data recovery mechanism can be set up to quickly restore data when the system fails to ensure the continuous operation of the system.

[0023] In order to facilitate management and maintenance, the data processor can also provide a real-time monitoring interface to display the current indoor and outdoor temperatures of each area. The real-time monitoring interface can be in the form of charts, tables, etc. to intuitively display the temperature change trend and current status.

[0024] In summary, the specific implementation process of indoor and outdoor temperature data collection includes: Sensor deployment: Select appropriate locations in different areas inside the building (such as living rooms, bedrooms, offices, etc.) and outside to install temperature sensors. Ensure that the sensors can accurately measure the temperature of each area while avoiding interference from other factors. Collection frequency setting: Set a reasonable data collection frequency according to the needs of the specific application scenario. Usually, the collection frequency of indoor temperature sensors and outdoor temperature sensors can be set to once every minute or every five minutes. Data transmission configuration: Select appropriate communication technology (such as Wi-Fi, Zigbee, LoRa, etc.), configure the data transmission parameters of the sensor, and ensure that the data can be stably transmitted to the data processor. For wired transmission, wiring and connection settings are required. Data storage management: Configure the local storage parameters of the sensor to ensure that the data is stored locally for a period of time to prevent data loss due to network failures. At the same time, configure the cloud storage parameters of the data processor to store the received data in the cloud server. Data verification and cleaning: Write data verification and cleaning programs to verify and clean the received data. Specific methods include checking whether the timestamps of the data are continuous, whether the temperature values ​​are within a reasonable range, and using statistical methods for anomaly detection and data cleaning. Data synchronization and backup: Set up data synchronization tasks to ensure that all sensor data can be synchronized to the data processor in a timely manner. Configure data backup strategies to regularly back up data stored in the cloud to prevent data loss or damage. Real-time monitoring interface: Develop a real-time monitoring interface to display the current indoor and outdoor temperatures of each area.

[0025] Through the above steps, the time queue of indoor temperature values ​​and the time queue of outdoor temperature values ​​collected by the temperature sensor can be efficiently and accurately received, providing reliable data support for subsequent environmental analysis.

[0026] In the above-mentioned building management system 100 for energy conservation and emission reduction, the humidity data acquisition module 120 is used to receive the time queue of indoor humidity values ​​and the time queue of outdoor humidity values ​​collected by the humidity sensor. The present application takes into account that humidity has a significant impact on people's thermal perception. Even if the temperature is the same, different humidity levels will make people feel different levels of comfort. For example, in a high humidity environment, even if the temperature is not high, people will feel stuffy and uncomfortable; while in a low humidity environment, even if the temperature is high, people may feel cooler. In addition, the humidity difference between indoors and outdoors will also affect the energy consumption of the air-conditioning system. Therefore, the present application further obtains indoor humidity data and outdoor humidity data to more accurately evaluate the impact of indoor and outdoor environments on the energy consumption of the air-conditioning system, improve the comfort of the occupants, and achieve more refined energy-saving control.

[0027] Similarly, the implementation process of humidity data collection is similar to that of the temperature data collection module, but it is optimized for the characteristics of humidity data. First, the deployment of humidity sensors needs to consider air circulation and the representativeness of ambient humidity. They are usually installed in locations with good air circulation and that can reflect the average humidity of the area, avoiding interference from other factors (such as vents, water vapor sources, etc.). Similarly, outside the building, humidity sensors should be placed in locations that can represent the humidity of the surrounding environment, avoiding direct influence from natural factors such as rain and fog.

[0028] Secondly, the frequency of humidity data collection needs to take into account the characteristics of humidity changes. Humidity changes are usually slower than temperature changes, so the collection frequency can be appropriately reduced. Usually, the collection frequency of indoor humidity sensors and outdoor humidity sensors can be set to once every ten minutes or every half an hour.

[0029] Humidity data verification and cleaning are equally important to ensure the accuracy and reliability of the data. During the data verification process, you can check whether the timestamps of the data are continuous and whether the readings of the humidity sensor are within the expected range. For example, indoor humidity is usually between 30% and 60%, while outdoor humidity may vary depending on geographical location and seasonal changes. If the reading is outside the normal range, it should be marked as abnormal data. Data cleaning involves eliminating abnormal readings due to sensor failure or other reasons. Statistical analysis methods, such as calculating the mean and standard deviation, can be used to identify and exclude outliers.

[0030] The data transmission, storage management and synchronous backup strategies of humidity data are similar to those of temperature data, but special attention should be paid to the characteristics of humidity data. For example, the synchronization frequency of humidity data can be slightly lower than that of temperature data to reduce unnecessary data transmission and storage pressure. The data backup strategy should ensure the integrity and consistency of the backup data. At the same time, the recovery test of the backup data should be performed regularly to verify the effectiveness of the backup. Through the above measures, the accurate collection and effective use of humidity data can be ensured, providing strong support for the comfort and energy-saving management of the building environment.

[0031] In the above-mentioned building management system 100 for energy conservation and emission reduction, the environmental difference periodicity analysis module 130 is used to perform periodic analysis of indoor and outdoor environmental differences based on the time queue of the indoor temperature value and the time queue of the outdoor temperature value, the time queue of the indoor humidity value and the time queue of the outdoor humidity value to obtain multiple indoor and outdoor temperature difference subsequences and multiple indoor and outdoor humidity difference subsequences. Figure 3 FIG. 1 is a block diagram of an environmental difference periodic analysis module in a building management system for energy saving and emission reduction according to an embodiment of the present application. Figure 3As shown, the environmental difference periodicity analysis module 130 includes: an indoor-outdoor difference calculation unit 131, which is used to calculate the positional differences between the time queue of the indoor temperature value and the time queue of the outdoor temperature value, the time queue of the indoor humidity value and the time queue of the outdoor humidity value, respectively, to obtain the time queue of the indoor-outdoor temperature difference value and the time queue of the indoor-outdoor humidity difference value; a periodic feature extraction unit 132, which is used to extract the periodic features in the time queue of the indoor-outdoor temperature difference value and the time queue of the indoor-outdoor humidity difference value by using a singular spectrum decomposition method to obtain the multiple indoor-outdoor temperature difference subsequences and the multiple indoor-outdoor humidity difference subsequences.

[0032] Specifically, the indoor-outdoor difference calculation unit 131 is used to calculate the position difference between the time queue of the indoor temperature value and the time queue of the outdoor temperature value, and the time queue of the indoor humidity value and the time queue of the outdoor humidity value, respectively, to obtain the time queue of the indoor-outdoor temperature difference and the time queue of the indoor-outdoor humidity difference. That is, in order to more intuitively understand the heat exchange and moisture migration inside and outside the building, the present application calculates the position difference between the indoor and outdoor temperature data and the position difference between the humidity data to quantitatively represent the heat and humidity exchange characteristics of the indoor and outdoor environment, thereby providing a more intuitive data basis for the energy-saving optimization control of the air conditioner.

[0033] Specifically, the periodic feature extraction unit 132 is used to extract the periodic features in the time queue of the indoor and outdoor temperature difference and the time queue of the indoor and outdoor humidity difference using a singular spectrum decomposition method to obtain the multiple indoor and outdoor temperature difference subsequences and the multiple indoor and outdoor humidity difference subsequences. It should be understood that, considering the complexity of the indoor and outdoor environment, the time series changes of the temperature difference and the humidity difference are often not a simple linear relationship, but contain a variety of periodic components. Therefore, in order to effectively extract the time series change characteristics of the indoor and outdoor temperature difference and the indoor and outdoor humidity difference, the present application uses a singular spectrum decomposition method to process the time queue of the indoor and outdoor temperature difference and the time queue of the indoor and outdoor humidity difference respectively, so as to decompose the complex time series data into several independent periodic components, generate multiple indoor and outdoor temperature difference subsequences and multiple indoor and outdoor humidity difference subsequences, thereby providing a more accurate analysis basis for the subsequent air conditioning temperature optimization control.

[0034] In a specific example of the present application, the periodic feature extraction unit is used to: construct a trajectory matrix of the time queue of the indoor and outdoor temperature difference based on a preset sliding window size to obtain an indoor and outdoor temperature difference time series matrix; perform singular value decomposition on the indoor and outdoor temperature difference time series matrix to obtain a sequence of left singular eigenvectors of the indoor and outdoor temperature difference and a sequence of right singular eigenvectors of the indoor and outdoor temperature difference; group the sequence of left singular eigenvectors of the indoor and outdoor temperature difference and the sequence of right singular eigenvectors of the indoor and outdoor temperature difference, and perform time queue reconstruction on each group of corresponding left singular eigenvectors of the indoor and outdoor temperature difference and right singular eigenvectors of the indoor and outdoor temperature difference to obtain the multiple indoor and outdoor temperature difference subsequences.

[0035] Specifically, taking the time queue of the indoor and outdoor temperature difference as an example, first, a preset sliding window is used to perform sliding sampling with a step size of 1 on the time queue of the indoor and outdoor temperature difference, and each sampling window is arranged into an indoor and outdoor temperature difference time series matrix, wherein each column of the indoor and outdoor temperature difference time series matrix is ​​a sampling window.

[0036] Next, the indoor and outdoor temperature difference time series matrix is ​​processed by singular value decomposition (SVD) technology to decompose it into a group of singular values ​​and left singular vectors and right singular vectors corresponding to each singular value, thereby obtaining a sequence of left singular eigenvectors of the indoor and outdoor temperature difference and a sequence of right singular eigenvectors of the indoor and outdoor temperature difference, so as to reveal the main dynamic change characteristics in the indoor and outdoor temperature difference time queue.

[0037] Then, according to the size of each singular value, the singular value and its corresponding left singular eigenvector of the indoor and outdoor temperature difference and right singular eigenvector of the indoor and outdoor temperature difference are grouped, and the singular values ​​and singular vectors of each group are reconstructed to generate a reconstructed time series matrix: ,in, is a matrix composed of the left singular eigenvectors of the indoor and outdoor temperature differences of the same group, is a matrix composed of the right singular eigenvectors of the indoor and outdoor temperature differences of the same group, is a diagonal matrix composed of the permutations of the singular values ​​of the same group, To reconstruct the time series sub-matrix of the indoor and outdoor temperature difference.

[0038] Finally, by calculating the column average of each group of reconstructed indoor and outdoor temperature difference time series submatrices, each reconstructed indoor and outdoor temperature difference time series submatrix is ​​converted back to the time queue form, thereby obtaining multiple indoor and outdoor temperature difference subsequences. Similarly, for the time queue of indoor and outdoor humidity difference, the above singular spectrum decomposition method is also used to obtain multiple indoor and outdoor humidity difference subsequences. In this way, the noise in the data can be effectively removed, and the main time series change trends and periodic components can be retained, thereby further improving the quality of the data and providing a more reliable basis for the energy-saving optimization control of the air-conditioning system.

[0039] In the above-mentioned building management system 100 for energy saving and emission reduction, the kernel feature extraction module 140 is used to perform kernel feature extraction based on multi-dimensional energy potential analysis on the multiple indoor and outdoor temperature difference subsequences and the multiple indoor and outdoor humidity difference subsequences to obtain indoor and outdoor temperature difference time series kernel feature vectors and indoor and outdoor humidity difference time series kernel feature vectors. That is, in order to further capture the core feature patterns of indoor and outdoor temperature difference and humidity difference time series changes, so as to better understand the dynamic change law of indoor and outdoor environment, so as to guide the energy-saving optimization control of air-conditioning system, the present application proposes a kernel feature extraction method, which performs multi-dimensional energy potential analysis on the multiple indoor and outdoor temperature difference subsequences and the multiple indoor and outdoor humidity difference subsequences respectively, so as to identify and strengthen the most representative and relevant feature patterns in multiple subsequences, eliminate noise and redundant information, and realize effective aggregation of kernel features, so as to generate indoor and outdoor temperature difference time series kernel feature vectors and indoor and outdoor humidity difference time series kernel feature vectors.

[0040] Figure 4 FIG. 1 is a block diagram of a kernel feature extraction module in a building management system for energy conservation and emission reduction according to an embodiment of the present application. Figure 4 As shown, the kernel feature extraction module 140 includes: a static energy potential analysis unit 141, which is used to perform static energy potential analysis on each indoor and outdoor temperature difference subsequence in the multiple indoor and outdoor temperature difference subsequences to obtain multiple indoor and outdoor temperature difference subsequences static energy potential factors; an initial aggregation unit 142, which is used to calculate the positional mean sequence of the multiple indoor and outdoor temperature difference subsequences to obtain the initial aggregation representation sequence of the indoor and outdoor temperature difference subsequences; a dynamic energy potential analysis unit 143, which is used to perform dynamic energy potential analysis on each indoor and outdoor temperature difference subsequence in the multiple indoor and outdoor temperature difference subsequences based on the initial aggregation representation sequence of the indoor and outdoor temperature difference subsequences to obtain multiple indoor and outdoor temperature difference subsequences dynamic energy potential factors; a kernel feature aggregation unit 144, which is used to perform kernel feature aggregation on the multiple indoor and outdoor temperature difference subsequences based on the multiple indoor and outdoor temperature difference subsequences static energy potential factors and the multiple indoor and outdoor temperature difference subsequences dynamic energy potential factors to obtain the indoor and outdoor temperature difference time series kernel feature vector.

[0041] Specifically, the static energy potential analysis unit 141 is used to: divide the fourth-order central moment of the indoor and outdoor temperature difference subsequence by the fourth power of its sequence variance as its static energy potential factor to obtain the static energy potential factor of the indoor and outdoor temperature difference subsequence, which is expressed by the formula: in, represents the multiple indoor and outdoor temperature difference subsequences, , , and respectively represent the first, second, and third indoor and outdoor temperature difference subsequences. and Indoor and outdoor temperature difference subsequences, is the number of the indoor and outdoor temperature difference subsequences, Indicates the The first indoor and outdoor temperature difference subsequence The value of the position, and Respectively represent the The sequence mean and sequence variance of the indoor and outdoor temperature difference subsequences, represents the length of the indoor and outdoor temperature difference subsequence, Indicates the The static energy potential factor of a subsequence of indoor and outdoor temperature differences.

[0042] Here, taking the multiple indoor and outdoor temperature difference subsequences as an example, first, for each indoor and outdoor temperature difference subsequence, by calculating its static energy potential factor, the information density it carries is revealed, and the inherent properties and information importance of each indoor and outdoor temperature difference subsequence when it is independent of other indoor and outdoor temperature difference subsequences are evaluated.

[0043] Specifically, the dynamic energy potential analysis unit 143 is used to: calculate the weighted sum of the indoor and outdoor temperature difference subsequence and the initial aggregation representation sequence of the indoor and outdoor temperature difference subsequence by position to obtain the dynamic energy potential representation sequence of the indoor and outdoor temperature difference subsequence; use a preset bias term to perform point addition modulation on each position in the dynamic energy potential representation sequence of the indoor and outdoor temperature difference subsequence, and calculate the square of the L2 norm of the dynamic energy potential representation sequence of the indoor and outdoor temperature difference subsequence after point addition modulation as its dynamic energy potential factor to obtain the dynamic energy potential factor of the indoor and outdoor temperature difference subsequence, which is expressed by the formula: in, is the initial aggregation representation sequence of the indoor and outdoor temperature difference subsequence, Indicates the The dynamic energy potential factor of the indoor and outdoor temperature difference subsequence, represents the bias term, and represents different weight parameters, represents dot product, It means calculating the square of L2 norm of the sequence.

[0044] That is, considering that each indoor and outdoor temperature difference subsequence represents a specific pattern of periodic temperature difference change characteristics, it has a certain contribution and influence on the overall dynamic change of the indoor and outdoor temperature difference. Therefore, the multiple indoor and outdoor temperature difference subsequences are further calculated by position mean to construct a typical example reflecting the overall time series characteristics of the indoor and outdoor temperature difference, and the overall time series change pattern of the indoor and outdoor temperature difference is captured, thereby obtaining the initial aggregation representation sequence of the indoor and outdoor temperature difference subsequences. Then, the dynamic energy potential factor of each indoor and outdoor temperature difference subsequence relative to the initial aggregation representation sequence of the indoor and outdoor temperature difference subsequence is calculated to capture the relative relationship between the data and measure its interaction and influence with the overall time series change pattern of the indoor and outdoor temperature difference.

[0045] Specifically, the kernel feature aggregation unit 144 includes: a potential superposition subunit, which is used to determine the superposition state potential factor of each indoor and outdoor temperature difference subsequence in the multiple indoor and outdoor temperature difference subsequences based on the static potential factors of the multiple indoor and outdoor temperature difference subsequences and the dynamic potential factors of the multiple indoor and outdoor temperature difference subsequences to obtain the superposition state potential factors of the multiple indoor and outdoor temperature difference subsequences; a weighting subunit, which is used to input the superposition state potential factors of the multiple indoor and outdoor temperature difference subsequences into the sigmoid function for normalization processing, and then perform mask inactivation processing on the normalized superposition state potential factors of the multiple indoor and outdoor temperature difference subsequences based on a preset mask threshold to obtain the characteristic potential weight coefficients of the multiple indoor and outdoor temperature difference subsequences; a weighted aggregation subunit, which is used to calculate the position-weighted sum of the multiple indoor and outdoor temperature difference subsequences based on the characteristic potential weight coefficients of the multiple indoor and outdoor temperature difference subsequences to obtain the indoor and outdoor temperature difference time series kernel feature vector. In a specific example of the present application, the energy potential superposition subunit is used to: use the squares of the static energy potential factor and the dynamic energy potential factor of the indoor and outdoor temperature difference subsequence as exponents, calculate the exponential function value with base e to obtain the exponential static energy potential factor and the exponential dynamic energy potential factor, and add the exponential static energy potential factor and the exponential dynamic energy potential factor to obtain the superposition state energy potential factor of the indoor and outdoor temperature difference subsequence.

[0046] The operation process of the kernel feature aggregation unit 144 can be expressed as follows: in, Indicates A normalized indoor and outdoor temperature difference subsequence superposition state energy potential factor, is the mask function, Indicates the preset mask threshold, For the The characteristic energy potential weight coefficient of the indoor and outdoor temperature difference subsequence, represents the indoor and outdoor temperature difference time series kernel feature vector.

[0047] That is, the static energy potential factor and dynamic energy potential factor of each indoor and outdoor temperature difference subsequence are further integrated to comprehensively consider their static and dynamic properties to form a superposition state energy potential factor, which is used as a comprehensive indicator to reflect the composite influence of each indoor and outdoor temperature difference subsequence in the overall data, and to more comprehensively evaluate its contribution to the core change pattern of the indoor and outdoor temperature difference. Then, based on the gated mask mechanism, each generated superposition state energy potential factor is normalized and weighted, so as to perform weighted aggregation on the multiple indoor and outdoor temperature difference subsequences, thereby explicitly marking the periodic component information that has a greater impact on the core change pattern of the indoor and outdoor temperature difference, filtering out irrelevant information, and obtaining the indoor and outdoor temperature difference time series kernel feature vector, thereby further enhancing the feature expression of the indoor and outdoor temperature difference time series data and improving the control accuracy of the air conditioning temperature.

[0048] In the above energy-saving and emission-reduction building management system 100, the air conditioning temperature optimization management module 150 is used to generate a management optimization instruction for the air conditioning temperature setting value based on the fusion characteristics of the indoor and outdoor temperature difference time series kernel feature vector and the indoor and outdoor humidity difference time series kernel feature vector. Figure 5 FIG. 1 is a block diagram of an air conditioning temperature optimization management module in a building management system for energy conservation and emission reduction according to an embodiment of the present application. Figure 5 As shown, the air conditioning temperature optimization management module 150 includes: a feature cascade unit 151, which is used to cascade the indoor and outdoor temperature difference timing kernel feature vector and the indoor and outdoor humidity difference timing kernel feature vector into an indoor and outdoor temperature-humidity timing kernel cascade feature vector; a decoding unit 152, which is used to input the indoor and outdoor temperature-humidity timing kernel cascade feature vector into a decoder-based building management module to obtain the management optimization instruction, and the management optimization instruction includes a decoded value of a recommended air conditioning temperature setting value.

[0049] That is, in order to more comprehensively describe the time series changes of the indoor and outdoor environments, the present application further cascades the indoor and outdoor temperature difference time series kernel feature vector and the indoor and outdoor humidity difference time series kernel feature vector into an indoor and outdoor temperature-humidity time series kernel cascade feature vector, so as to comprehensively consider the joint impact of temperature and humidity on air-conditioning energy consumption, and provide more comprehensive decision support for the control of air-conditioning temperature.

[0050] Finally, the building management module is constructed through the decoder model, and the indoor and outdoor temperature-humidity time series kernel cascade feature vector is decoded and regressed to generate a decoded value of the recommended air-conditioning temperature setting value. In the technical solution of the present application, the decoder adopts a multi-layer neural network architecture, and through deep learning and feature extraction of the indoor and outdoor temperature-humidity time series kernel cascade feature vector, it captures the deep-level laws and patterns of indoor and outdoor environmental changes, and maps them to the decoding space of the air-conditioning temperature setting value, thereby generating a recommended air-conditioning temperature setting value for the current indoor and outdoor environmental conditions. In this way, the air-conditioning system can automatically adjust the operating status according to real-time indoor and outdoor environmental data to achieve the dual goals of energy saving and comfort.

[0051] Preferably, before the indoor and outdoor temperature-humidity timing series kernel cascade feature vector is input into a decoder-based building management module to obtain management optimization instructions, the indoor and outdoor temperature-humidity timing series kernel cascade feature vector is subjected to feature joint distribution optimization to obtain an optimized indoor and outdoor temperature-humidity timing series kernel cascade feature vector.

[0052] Specifically, the process of optimizing the joint distribution of features includes the following steps: Based on the eigenvalues ​​of the indoor and outdoor temperature-humidity time series kernel cascade feature vector Distance and Distance to get indoor and outdoor temperature-humidity time series kernel cascade-distance matrix and indoor and outdoor temperature-humidity time series kernel cascade two distance matrices ; Calculate the weighted sum of the indoor and outdoor temperature-humidity time series kernel cascade one distance matrix and the indoor and outdoor temperature-humidity time series kernel cascade two distance matrix to obtain the indoor and outdoor temperature-humidity time series kernel cascade joint distance matrix ,in, and represents the weighted hyperparameter, It means adding by position. It means point multiplication by position; Determine the eigenvalues ​​of the indoor and outdoor temperature-humidity time series kernel cascade joint distance matrix , and form the indoor and outdoor temperature-humidity time series kernel cascade joint distance eigenvector ; The indoor and outdoor temperature-humidity time series kernel cascade feature vector as a row vector is matrix-multiplied with the indoor and outdoor temperature-humidity time series kernel cascade-distance matrix to obtain the indoor and outdoor temperature-humidity time series kernel cascade-distance query vector ,in, represents the indoor and outdoor temperature-humidity time series kernel cascade feature vector, Represents matrix multiplication; The indoor and outdoor temperature-humidity time series kernel cascade two distance matrix is ​​matrix-multiplied with the indoor and outdoor temperature-humidity time series kernel cascade feature vector autocorrelation matrix to obtain the indoor and outdoor temperature-humidity time series kernel cascade two distance correlation matrix ,in, Represents the transpose of a vector; After matrix multiplication of the indoor and outdoor temperature-humidity time series kernel cascade one distance query vector and the indoor and outdoor temperature-humidity time series kernel cascade two distance association matrix, further dot multiplication with the indoor and outdoor temperature-humidity time series kernel cascade joint distance eigenvector is performed to obtain the optimized indoor and outdoor temperature-humidity time series kernel cascade feature vector Finally, the optimized indoor and outdoor temperature-humidity timing kernel cascade feature vector is input into the decoder-based building management module to obtain management optimization instructions.

[0053] Therefore, the one distance matrix and the two distance matrices of the indoor and outdoor temperature-humidity time series kernel cascade feature vector are used as the fine-grained metric association cluster representation of the indoor and outdoor temperature-humidity time series kernel cascade feature vector, and the dynamic programming of the relationship between association clusters of different association clusters is performed on the indoor and outdoor temperature-humidity time series kernel cascade feature vector and the self-association representation of the indoor and outdoor temperature-humidity time series kernel cascade feature vector respectively to simulate the sparse activation based on neuron clusters of the association system, and the fine-grained predictable sparsity of the indoor and outdoor temperature-humidity time series kernel cascade feature vector is coordinated with the intrinsic representation of the metric association cluster of the one distance matrix and the two distance matrices of the indoor and outdoor temperature-humidity time series kernel cascade feature vector. In this way, the significant inherent modal information in the source domain features can be more effectively retained, and the interference components in the feature distribution can be effectively masked, so as to improve the adversarial robustness of the manifold expression of the feature vector. In this way, the accuracy of the management optimization instructions obtained by the building management module based on the decoder input is improved.

[0054] In summary, the energy-saving and emission-reducing building management system according to the embodiment of the present application is explained, which uses artificial intelligence technology based on deep learning to perform time series analysis on indoor and outdoor temperature data and humidity data, respectively extracts the periodic change characteristics of the indoor and outdoor temperature difference and humidity difference, and further refines the kernel features to dig out the core feature patterns of the time series changes of the indoor and outdoor temperature difference and humidity difference, so as to realize intelligent recommendation and optimization control of the air-conditioning temperature setting value in the building on this basis. In this way, by monitoring the indoor and outdoor temperature and humidity changes in real time to automatically adjust the working state of the air conditioner, the comfort of the indoor environment can be effectively maintained and energy consumption can be reduced, thereby achieving the purpose of energy saving and emission reduction.

[0055] The basic principle of the present invention is described above in conjunction with specific embodiments. However, it should be pointed out that the advantages, strengths, effects, etc. mentioned in the present invention are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. must be possessed by each embodiment of the present invention. In addition, the specific details of the above embodiments are only for the purpose of illustration and facilitation of understanding, rather than limitation, and the above details do not limit the present invention to being implemented by adopting the above specific details.

Claims

1. A building management system for energy saving and emission reduction, characterized in that: include: A temperature data acquisition module, used for receiving a time queue of indoor temperature values ​​and a time queue of outdoor temperature values ​​collected by a temperature sensor; A humidity data acquisition module, used for receiving a time queue of indoor humidity values ​​and a time queue of outdoor humidity values ​​collected by a humidity sensor; An environmental difference periodicity analysis module, configured to perform periodicity analysis of indoor and outdoor environmental differences based on the time queue of the indoor temperature value and the time queue of the outdoor temperature value, the time queue of the indoor humidity value and the time queue of the outdoor humidity value to obtain a plurality of indoor and outdoor temperature difference subsequences and a plurality of indoor and outdoor humidity difference subsequences; A kernel feature extraction module, used for performing kernel feature extraction based on multidimensional energy potential analysis on the multiple indoor and outdoor temperature difference subsequences and the multiple indoor and outdoor humidity difference subsequences to obtain indoor and outdoor temperature difference time series kernel feature vectors and indoor and outdoor humidity difference time series kernel feature vectors; The air conditioning temperature optimization management module is used to generate a management optimization instruction for the air conditioning temperature setting value based on the fusion characteristics of the indoor and outdoor temperature difference time series kernel feature vector and the indoor and outdoor humidity difference time series kernel feature vector.

2. The building management system for energy saving and emission reduction according to claim 1 is characterized in that: The environmental difference periodic analysis module includes: An indoor-outdoor difference calculation unit, used to calculate the position difference between the time queue of the indoor temperature value and the time queue of the outdoor temperature value, and the time queue of the indoor humidity value and the time queue of the outdoor humidity value, respectively, to obtain a time queue of the indoor-outdoor temperature difference value and a time queue of the indoor-outdoor humidity difference value; The periodic feature extraction unit is used to extract the periodic features in the time queue of the indoor and outdoor temperature difference and the time queue of the indoor and outdoor humidity difference by using a singular spectrum decomposition method to obtain the multiple indoor and outdoor temperature difference subsequences and the multiple indoor and outdoor humidity difference subsequences.

3. The building management system for energy saving and emission reduction according to claim 2 is characterized in that: The periodic feature extraction unit is used to: Constructing a trajectory matrix of the time queue of the indoor and outdoor temperature difference based on a preset sliding window size to obtain a time series matrix of the indoor and outdoor temperature difference; Performing singular value decomposition on the indoor and outdoor temperature difference time series matrix to obtain a sequence of left singular eigenvectors of the indoor and outdoor temperature difference and a sequence of right singular eigenvectors of the indoor and outdoor temperature difference; The sequence of the left singular eigenvector of the indoor and outdoor temperature difference and the sequence of the right singular eigenvector of the indoor and outdoor temperature difference are grouped, and the time queue reconstruction is performed on each group of corresponding left singular eigenvectors of the indoor and outdoor temperature difference and right singular eigenvectors of the indoor and outdoor temperature difference to obtain the multiple indoor and outdoor temperature difference subsequences.

4. The building management system for energy saving and emission reduction according to claim 3 is characterized in that: The kernel feature extraction module comprises: A static energy potential analysis unit, used for performing a static energy potential analysis on each of the plurality of indoor and outdoor temperature difference subsequences to obtain a plurality of indoor and outdoor temperature difference subsequence static energy potential factors; An initial aggregation unit, used for calculating the position mean sequence of the plurality of indoor and outdoor temperature difference subsequences to obtain an initial aggregation representation sequence of the indoor and outdoor temperature difference subsequences; A dynamic energy potential analysis unit, configured to perform a dynamic energy potential analysis on each of the plurality of indoor and outdoor temperature difference subsequences based on an initial aggregation representation sequence of the indoor and outdoor temperature difference subsequences to obtain a plurality of dynamic energy potential factors of the indoor and outdoor temperature difference subsequences; The kernel feature aggregation unit is used to perform kernel feature aggregation on the multiple indoor and outdoor temperature difference subsequences based on the multiple indoor and outdoor temperature difference subsequences static energy potential factors and the multiple indoor and outdoor temperature difference subsequences dynamic energy potential factors to obtain the indoor and outdoor temperature difference time series kernel feature vector.

5. The building management system for energy saving and emission reduction according to claim 4 is characterized in that: The static energy potential analysis unit is used for: The calculation result of dividing the fourth-order central moment of the indoor-outdoor temperature difference subsequence by the fourth power of its sequence variance is used as its static energy potential factor to obtain the static energy potential factor of the indoor-outdoor temperature difference subsequence.

6. The building management system for energy saving and emission reduction according to claim 5 is characterized in that: The dynamic energy potential analysis unit is used to: Calculating the position-weighted sum of the indoor-outdoor temperature difference subsequence and the initial aggregation representation sequence of the indoor-outdoor temperature difference subsequence to obtain a dynamic energy potential representation sequence of the indoor-outdoor temperature difference subsequence; After using a preset bias item to perform point-adding modulation on each position in the dynamic energy potential representation sequence of the indoor and outdoor temperature difference subsequence, the square of the L2 norm of the dynamic energy potential representation sequence of the indoor and outdoor temperature difference subsequence after point-adding modulation is calculated as its dynamic energy potential factor to obtain the dynamic energy potential factor of the indoor and outdoor temperature difference subsequence.

7. The building management system for energy saving and emission reduction according to claim 6 is characterized in that: The kernel feature aggregation unit comprises: An energy potential superposition subunit is used to determine the superposition state energy potential factor of each indoor and outdoor temperature difference subsequence in the multiple indoor and outdoor temperature difference subsequences based on the multiple indoor and outdoor temperature difference subsequences static energy potential factors and the multiple indoor and outdoor temperature difference subsequences dynamic energy potential factors to obtain multiple indoor and outdoor temperature difference subsequences superposition state energy potential factors; A weighting subunit is used to input the multiple indoor and outdoor temperature difference subsequence superposition state energy potential factors into a sigmoid function for normalization processing, and then perform mask inactivation processing on the normalized multiple indoor and outdoor temperature difference subsequence superposition state energy potential factors based on a preset mask threshold to obtain multiple indoor and outdoor temperature difference subsequence characteristic energy potential weight coefficients; The weighted aggregation subunit is used to calculate the position-weighted sum of the multiple indoor and outdoor temperature difference subsequences based on the characteristic energy potential weight coefficients of the multiple indoor and outdoor temperature difference subsequences to obtain the indoor and outdoor temperature difference time series kernel feature vector.

8. The building management system for energy saving and emission reduction according to claim 7 is characterized in that: The energy potential superposition subunit is used for: The squares of the static energy potential factor and the dynamic energy potential factor of the indoor and outdoor temperature difference subsequence are respectively used as exponents, and the exponential function values ​​with base e are calculated to obtain the exponential static energy potential factor and the exponential dynamic energy potential factor, and the exponential static energy potential factor and the exponential dynamic energy potential factor are added and summed to obtain the superposition state energy potential factor of the indoor and outdoor temperature difference subsequence.

9. The building management system for energy saving and emission reduction according to claim 8, characterized in that: The air conditioning temperature optimization management module includes: A feature cascading unit, used for cascading the indoor and outdoor temperature difference time series kernel feature vector and the indoor and outdoor humidity difference time series kernel feature vector into an indoor and outdoor temperature-humidity time series kernel cascade feature vector; A decoding unit is used to input the indoor and outdoor temperature-humidity time series kernel cascade feature vector into a decoder-based building management module to obtain the management optimization instruction, wherein the management optimization instruction includes a decoded value of a recommended air conditioning temperature setting value.

Images