Operation period building energy-saving low-carbon control method and system based on artificial intelligence

Abstract

The invention relates to the technical field of self-adaptive control, in particular to an operation period building energy-saving low-carbon control method and system based on artificial intelligence, and the method comprises the following steps: collecting and analyzing the room use condition and illumination intensity, and obtaining an activity type recognition result; according to the invention, based on real-time analysis of room use conditions and illumination intensity, use modes of different areas are identified, and in combination with natural light changes and indoor activity requirements, required illumination intensity and color temperature adjustment parameters are dynamically calculated, so that accurate matching of illumination environments is realized. A dynamic feedback mechanism is established by monitoring the adjustment state of the lighting equipment in real time in combination with the change of the room use mode, the lighting deviation is adaptively corrected, and energy waste is avoided. By introducing reinforcement learning, the ventilation strategy can be continuously optimized through historical data and real-time feedback, personalized air quality management is achieved, and single adjustment is not executed based on a fixed threshold value.

Description

Technical Field

[0001] The present invention relates to the field of adaptive control technology, and in particular to an artificial intelligence-based building energy-saving and low-carbon control method and system during operation. Background Art

[0002] Adaptive control is a technology that can adjust control strategies in real time according to environmental changes and system status, and is widely used in the automatic regulation of complex dynamic systems. The energy-saving and low-carbon control method for buildings during operation aims to intelligently regulate the energy consumption and carbon emissions of buildings during operation.

[0003] In the existing technology, lighting control is usually based on preset brightness values ​​or fixed switch logic, and fails to dynamically calculate lighting needs in combination with space usage characteristics and natural light intensity, which may lead to excessive or low illumination in some areas, affecting the rationality of energy management. At the same time, the ventilation system is mostly regulated by passive ventilation after the air quality reaches the set threshold, and fails to make predictive adjustments to environmental trends. When the air quality changes rapidly, there may be problems such as adjustment lag or frequent start and stop, which increases energy consumption. Therefore, improvements are needed. Summary of the invention

[0004] The purpose of the present invention is to solve the shortcomings of the prior art and to propose an energy-saving and low-carbon control method and system for buildings in operation period based on artificial intelligence.

[0005] In order to achieve the above purpose, the present invention adopts the following technical solution, an artificial intelligence-based building energy-saving and low-carbon control method during operation period, comprising the following steps:

[0006] Collect and analyze room usage and light intensity to obtain activity type recognition results; calculate required lighting intensity and color temperature adjustment parameters based on the natural light intensity and indoor activity requirements in the activity type recognition results, and generate lighting adjustment requirements;

[0007] Based on the lighting adjustment requirements, the brightness and color temperature of the LED lamps are adjusted in real time to generate lamp adjustment results; the lamp adjustment results are compared with the room usage pattern to identify the deviation of the lighting pattern and generate lighting pattern optimization instructions;

[0008] Monitor and analyze the environmental quality according to the indoor and outdoor air quality and temperature and humidity data to obtain the environmental quality analysis results; adjust the ventilation volume and ventilation mode of the ventilation equipment according to the environmental quality analysis results to generate ventilation adjustment results;

[0009] Based on the ventilation adjustment results, reinforcement learning is used to analyze the deviation between the adjustment effect and the expected target, and a ventilation optimization strategy is generated.

[0010] Preferably, the steps of obtaining the activity type identification result are:

[0011] Collect real-time light intensity and room usage data, including light intensity, room entry and exit frequency and residence time, to generate environmental monitoring data;

[0012] Based on the environmental monitoring data, machine learning is used to learn and analyze the rules of activities in the room to obtain behavior pattern analysis results;

[0013] According to the behavior pattern analysis result, the activity type in the room is identified and verified, and the activity type includes rest, work or meeting, and the activity type identification result is obtained.

[0014] Preferably, the steps of obtaining the lighting adjustment requirement are:

[0015] Extract the natural light intensity information from the activity type recognition result, and analyze the light distribution of the room in different time periods in combination with the indoor activity needs, obtain the light change trend of each area and the indoor use scene characteristics, and obtain the lighting demand data;

[0016] According to the lighting demand data, the required lighting intensity is calculated using the following formula:

[0017]

[0018] in, is the required lighting intensity, For the The intensity of indoor activities For the The natural light intensity at For the The current lighting brightness at is the average lighting brightness of the current environment, For the The wall light absorption coefficient at For the The light reflection coefficient of the furniture surface at The total number of positions used for calculating lighting intensity. is the position number to which the wall absorption coefficient is referred, The total number of furniture surfaces referenced;

[0019] According to the lighting demand data, the required color temperature adjustment parameters are calculated, and the calculation formula is:

[0020]

[0021] in, Adjust the parameters for the desired color temperature, For the The current light source brightness at For the The current color temperature value at For the The window transmittance at For the The light source environment interference factor at is the preset target color temperature, is the color temperature measurement value of the current environment, The total number of light source points used as reference when calculating the color temperature adjustment parameters. Total number of references for windows and environmental distractions;

[0022] Based on the lighting intensity and color temperature adjustment parameters, the matching relationship between the lighting demand and the activity type is analyzed, the data points with abnormal fluctuations in the calculated values ​​are corrected, the lighting intensity and color temperature adjustment parameters are determined, and the lighting adjustment demand is generated.

[0023] Preferably, the steps of obtaining the lamp adjustment result are:

[0024] Based on the lighting adjustment requirement, extract the target lighting intensity and target color temperature parameters in the lighting adjustment requirement, analyze the brightness and color temperature setting values ​​of the current LED lamp, calculate the difference between the two, analyze whether it is necessary to adjust the brightness, color temperature separately or the brightness and color temperature at the same time, and obtain the lighting adjustment instruction;

[0025] According to the lighting adjustment instruction, the control module of the LED lamp is called to send brightness and color temperature adjustment signals to the target lamp to form a lamp adjustment result.

[0026] Preferably, the steps of obtaining the lighting mode optimization instruction are:

[0027] According to the lamp adjustment result, the lighting mode deviation value is calculated, and the calculation formula is:

[0028]

[0029] in, is the lighting mode deviation value, is the brightness value of the lamp in the current period, is the target brightness value, Adjust the impact factor for brightness, is the color temperature value of the current lamp, is the target color temperature value, is the ambient light source interference factor;

[0030] Based on the lighting mode deviation value, the rationality of the lighting mode adjustment is analyzed, and it is determined whether the brightness and color temperature deviations of each lamp exceed the set range, and a lighting mode optimization instruction is generated.

[0031] Preferably, the steps for obtaining the environmental quality analysis results are:

[0032] Collect indoor and outdoor air quality and temperature and humidity data, integrate the data, and obtain pre-processed environmental data;

[0033] Based on the pre-processed environmental data, the comprehensive environmental quality index is calculated using the following formula:

[0034]

[0035] in, is the comprehensive index of environmental quality, is the indoor carbon dioxide concentration, is the outdoor ozone concentration, is the indoor humidity, is the comfortable humidity threshold for human body, is the indoor temperature, is the outdoor temperature, is the indoor PM2.5 concentration, is the outdoor PM10 concentration, is the wind speed;

[0036] Based on the comprehensive environmental quality index, the state of the overall environmental quality is evaluated, time periods with abnormal environmental quality are screened, and environmental quality analysis results are generated.

[0037] Preferably, the steps of obtaining the ventilation adjustment result are:

[0038] According to the environmental quality analysis result, by comparing with the predetermined health threshold, the ventilation parameters that need to be adjusted are identified to form ventilation adjustment requirements;

[0039] Using the ventilation control interface, adjusting the ventilation volume and ventilation mode according to the ventilation adjustment requirements, including increasing or decreasing the wind speed and switching from the circulation mode to the external ventilation mode;

[0040] Monitor the adjusted ventilation in real time, evaluate whether the effect of ventilation adjustment meets health and safety standards, and generate ventilation adjustment results.

[0041] Preferably, the steps for obtaining the ventilation optimization strategy are:

[0042] Inputting the ventilation adjustment result into the reinforcement learning model, learning and matching the adjustment result with the set environmental target by comparing the adjustment result, and generating an optimized learning result;

[0043] According to the optimized learning results, the ventilation settings are reconfigured to generate a ventilation optimization strategy.

[0044] The present invention provides an energy-saving and low-carbon control system, comprising:

[0045] Environmental monitoring module, which collects indoor and outdoor air quality, temperature and humidity data and generates environmental quality reports;

[0046] The lighting adjustment module calculates the required lighting intensity and color temperature adjustment parameters based on the natural light intensity and room activity requirements in the environmental quality report, and generates lighting adjustment requirements;

[0047] The ventilation control module adjusts the ventilation volume and ventilation mode of the ventilation equipment according to the environmental quality report and generates ventilation adjustment results;

[0048] The lamp adjustment module adjusts the brightness and color temperature of LED lamps based on lighting adjustment requirements, generates lamp adjustment results, compares the lamp adjustment results with the room usage pattern, identifies the deviation of the lighting pattern, and generates lighting pattern optimization instructions;

[0049] The optimization strategy module uses the ventilation adjustment results to perform reinforcement learning analysis, compares the deviation between the adjustment effect and the expected target, and generates a ventilation optimization strategy.

[0050] Compared with the prior art, the advantages and positive effects of the present invention are:

[0051] In the present invention, based on the real-time analysis of room usage and light intensity, the usage patterns of different areas are identified, and combined with the changes in natural light and the needs of indoor activities, the required lighting intensity and color temperature adjustment parameters are dynamically calculated to achieve accurate matching of the lighting environment. By real-time monitoring of the adjustment status of lighting equipment and combining the changes in room usage patterns, a dynamic feedback mechanism is established to adaptively correct lighting deviations to avoid energy waste. The introduction of reinforcement learning enables ventilation strategies to be continuously optimized through historical data and real-time feedback to achieve personalized air quality management, rather than performing a single adjustment based on a fixed threshold. Multiple systems are linked, and through the coordinated adjustment of multiple subsystems such as lighting and ventilation, the internal environment of the building can achieve optimal comfort with minimal energy consumption. For long-term operating buildings, energy consumption optimization is not limited to short-term adjustments, but through continuous learning and continuous adjustment of strategies, the system adapts to different usage scenarios and dynamic changes in the external environment to improve overall energy-saving efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a schematic diagram of the steps of the present invention. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0054] See also Figure 1 The present invention provides a technical solution, an artificial intelligence-based building energy-saving and low-carbon control method during operation, comprising the following steps:

[0055] Collect and analyze room usage and light intensity to obtain activity type recognition results; calculate required lighting intensity and color temperature adjustment parameters based on the natural light intensity and indoor activity requirements in the activity type recognition results, and generate lighting adjustment requirements;

[0056] Based on lighting adjustment requirements, the brightness and color temperature of LED lamps are adjusted in real time to generate lamp adjustment results; the lamp adjustment results are compared with the room usage pattern to identify the deviation of the lighting pattern and generate lighting pattern optimization instructions;

[0057] Monitor and analyze environmental quality based on indoor and outdoor air quality and temperature and humidity data to obtain environmental quality analysis results; adjust the ventilation volume and ventilation mode of ventilation equipment based on the environmental quality analysis results to generate ventilation adjustment results;

[0058] Based on the ventilation adjustment results, reinforcement learning is used to analyze the deviation between the adjustment effect and the expected target, and a ventilation optimization strategy is generated.

[0059] The steps to obtain the activity type recognition results are:

[0060] Collect real-time light intensity and room usage data, including light intensity, room entry and exit frequency and residence time, to generate environmental monitoring data;

[0061] Based on environmental monitoring data, machine learning is used to learn and analyze the rules of activities in the room to obtain behavioral pattern analysis results;

[0062] According to the behavior pattern analysis results, the activity types in the room are identified and verified, and the activity types include rest, work or meeting, and the activity type recognition results are obtained.

[0063] Specifically, referring to the original records provided by multiple light sensors and cameras, the light intensity and personnel entry and exit information corresponding to the current moment are extracted and the sampling moment is marked. In combination with the number of entries and exits and the length of stay of personnel in the same time period, the light intensity is compared with the preset interval, for example, the range of 0lx to 2000lx is compared. If the light intensity value exceeds 2000lx or is lower than 0lx, it is deemed to be inconsistent with the conventional range of indoor measurement and recorded as abnormal data. In combination with the frequency of personnel entry and exit, an additional reference interval of 0 to 10 times per minute can be set. If it exceeds this range, it is judged that the room is abnormally intensively used or abnormally vacant. In order to define the reference interval value, a maximum acceptable value or a minimum acceptable value is obtained by using the method based on the statistics of the past week plus two standard deviations. The valid values ​​of the lighting data and personnel information are grouped together for time series sorting and attached with timestamps. The personnel residence time is obtained by tracking the continuous time period of each person in the room. If the residence time falls between 0 minutes and 600 minutes, it is included in the normal range and no additional exclusion operation is performed. If it exceeds 600 minutes, the data is considered to be extreme and is placed separately in the abnormal group. Finally, the above-mentioned screened and marked content is integrated to obtain the environmental monitoring data.

[0064] Based on the environmental monitoring data obtained above, the light intensity sequence in each time period and the corresponding frequency of people entering and exiting and the length of stay are used as input samples. The pre-set supervised machine learning model is used for training. All samples are first divided into training sets and test sets, and the two are ensured to maintain a consistent proportion in the distribution of light and personnel values. Then, about 100 cycles are executed in batch iteration. Each iteration calculates the loss function based on the prediction results of the current model and the known label values. If the loss is greater than the specified threshold, the parameter update operation is performed. The specified threshold here is gradually reduced from the initial setting of 0.05 to 0.01. The parameter update adopts the stochastic gradient descent method and evaluates the accuracy of the model on the test set after each batch. If the accuracy is lower than 60% for a long time, the learning rate is adjusted from 0.01 to 0.001 and the cycle is re-executed. After each iteration, the current parameter state is saved to compare the final stable convergence. Finally, through multiple trainings, it is confirmed that the accuracy threshold of 80% is met and the iteration is stopped to obtain the behavior pattern analysis results.

[0065] According to the behavior pattern analysis results obtained above, the room activity feature vectors output in the analysis phase are compared one by one, and the pre-determined multi-classification judgment algorithm is used to perform classification judgment on the three labels of rest, work, and meeting and verify their classification confidence. Each record will first be compared with a classification threshold table when entering the judgment algorithm. The threshold table contains information such as the statistical median of light intensity corresponding to the three activity types, the amplitude of changes in personnel entering and exiting, and the average length of stay. The setting method of the threshold table is based on the sampling frequency and the historical mean calculation. For example, 100lx, 300lx, and 500lx are selected for light intensity. Three representative values ​​are used to set three reference ranges of 1, 3 and 6 times per minute for the amplitude of changes in personnel entry and exit, and 10, 60 and 180 minutes are defined for the average stay time. The activity feature vector is compared with the above thresholds one by one. If the corresponding interval of light intensity falls around 300lx, the amplitude of changes in personnel entry and exit falls around 3 times per minute and the average stay time is around 60 minutes, it is judged as a work type. If the corresponding interval values ​​match other combinations respectively, it is judged as a rest or meeting type. After all samples are judged, the corresponding activity type is output and verified and compared to obtain the activity type recognition result.

[0066] The steps to obtain lighting adjustment requirements are:

[0067] Extract the natural light intensity information from the activity type recognition results, and analyze the light distribution of the room at different time periods in combination with the indoor activity needs, obtain the light change trend of each area and the characteristics of the indoor use scene, and obtain the lighting demand data;

[0068] Based on the natural light intensity information contained in the activity type recognition results obtained above, first read and disassemble the timestamp record and the corresponding light value in the information, then extract the usage expectations for different time periods (such as 6:00 to 8:00 in the morning, 12:00 to 14:00 in the afternoon, and 18:00 to 22:00 in the evening) from the indoor activity demand list, clarify the preferred range of space brightness for each time period, and by comparing the light value with the reference interval of 0lx to 2000lx, determine which time periods may have high or low light intensity. For example, in the morning when the crowd has not yet entered the room on a large scale, the light intensity is often within a few hundred lx. If it exceeds 2000lx, it is regarded as an abnormally high value and is independently marked. To set the reference interval, it is necessary to combine the daily light levels measured in this room in the past month for statistics, calculate the light mean and maximum peak value, and check whether there are extreme deviations in the observation data. If extreme values ​​are found, a correction interval is obtained by removing the highest 1% and the lowest 1%, and then the corrected 0lx to 2000lx is used as the acceptable interval for the current room. The conventional lighting range is then analyzed, and the changes in lighting distribution in the room during adjacent time periods are analyzed. If it is detected that a certain area has rapid increases and decreases in lighting multiple times in a day (for example, from 200lx to 800lx in one minute), it is recorded that the area has a large fluctuation in the corresponding time period. Then, combined with the brightness settings for leisure, work or other high-precision demand scenes in the indoor activity demand list, the lighting fluctuation situation is compared with the threshold value listed in the demand. If a scene requires a lighting range of 300lx to 600lx and the measured value is maintained at 800lx, it is considered that the lighting in this period is too high and the corresponding time period index is recorded. Through this comparison of time periods and values, the lighting change trend in the room is further sorted out. Finally, the lighting records distributed in each area are mapped with the room usage scene characteristics, and the information included in the scene characteristics, such as activity duration, number of people, contrast requirements, etc., are matched one by one with the extracted lighting fluctuation data to form multiple groups of comparison arrays, and then the time periods and area numbers that meet the corresponding requirements are counted, and finally integrated into lighting demand data.

[0069] According to the lighting demand data, the required lighting intensity is calculated using the following formula:

[0070]

[0071] in, is the required lighting intensity, For the The intensity of indoor activities For the The natural light intensity at For the The current lighting brightness at is the average lighting brightness of the current environment, For the The wall light absorption coefficient at For the The light reflection coefficient of the furniture surface at The total number of positions used for calculating lighting intensity. is the position number to which the wall absorption coefficient is referred, The total number of furniture surfaces referenced;

[0072] The benefit of the formula is that it uses multiple factors, such as indoor activity intensity, natural light illumination, current lighting brightness and average brightness, wall light absorption coefficient and furniture reflection coefficient, to comprehensively calculate the required lighting intensity, thereby more carefully measuring brightness requirements in different locations and environments. By incorporating the optical properties of walls and furniture into the calculation, a lighting intensity assessment that is closer to actual indoor conditions is achieved.

[0073] The steps to obtain the parameters refer to the first The indoor activity intensity at a location is generally quantified by observing indicators such as the frequency of human movements in the room, the number of devices used, and the overall behavioral energy consumption. Infrared sensors and motion capture cameras can be installed at each location to record the amount of activity per unit time. If the amount of activity is high, a larger value can be assigned. For example, in the daily office area, if a large amount of walking or multiple device interactions are detected per unit time, the activity intensity can be set to 6, and if fewer movements are detected, it can be set to 2. An average activity intensity value can be obtained through actual monitoring for 7 consecutive days, and the average value is used as the benchmark for recording at each location. When an example measurement is performed on a certain location, if it is found that there are 30 people passing through per hour and the average stay time of each person is about 5 minutes, the activity intensity of this location is set to 4 based on experience.

[0074] The steps to obtain the parameters are as follows: The natural light illuminance at the place is recorded by the light sensor placed near the window or ceiling, and then converted to the indoor light. The window position with relatively consistent transparency can be selected to eliminate the influence of local shading. For a south-facing glass window, the outdoor light illuminance at 10 a.m. is 30,000 lx through calibration. Combined with the window's transmittance of about 30%, the indoor light illuminance can be 9,000 lx. According to the room's internal shading and the actual distance from the window, for example, if the distance to the window is about 3 meters, the light attenuation is about 25%, which can be converted to about 6,750 lx. The numerical value of .

[0075] The steps to obtain the parameters are as follows: The current lighting brightness at a certain location can be measured in real time by placing a illuminance meter 1 meter below the lamp. If a set of lighting equipment consists of multiple lamp heads, the brightness contributed by each lamp head is measured one by one and smoothed in the horizontal range, and the average brightness in the space is summarized to that location. When only one LED lamp is turned on at a certain location at 19:00 in the evening, the lighting brightness is about 400lx, then .

[0076] The steps for obtaining the parameter are as follows: this parameter represents the average lighting brightness of the current environment. Several illuminance meters can be arranged in the room and the brightness values ​​can be read synchronously and regularly. The average brightness can be obtained by adding up all these values ​​and dividing by the number of measuring points. Suppose there are 6 measuring points evenly distributed in a room, and the brightness data recorded at each measuring point are 380lx, 400lx, 410lx, 405lx, 390lx, and 395lx respectively. The sum is 2380lx, and then divided by 6 to get about 396.7lx. 396.7 is rounded off to get 397lx to represent the average lighting brightness of the current environment.

[0077] The steps to obtain the parameters are as follows: The light absorption coefficient of the wall at the location needs to be obtained through field testing or by consulting the material data manual. If the wall surface is painted with dark paint or has a matte texture, the absorption coefficient is usually larger. In the experiment, a light source of known intensity can be used to illuminate the wall surface, and the ratio of the incident light intensity to the reflected light intensity can be measured to determine the absorption rate. The absorption coefficient is obtained by subtracting the reflectivity from 1. If the reflectivity is 0.7, the absorption coefficient is 0.3. The specific value can be recorded in For a lounge, if the wall material is brighter and the reflectivity is 0.6, the absorption coefficient is 0.4.

[0078] The steps to obtain the parameters are as follows: The light reflectance coefficient of the furniture surface at the position is used to measure the degree of light reflection of the furniture surface. If there are multiple pieces of furniture in the room, the light reflectance of the surface material of each piece of furniture is measured separately. For example, the reflectance of a leather sofa is about 0.2 to 0.3, and the reflectance of a dark wooden tabletop may be around 0.1. By summarizing the reflectance values ​​of each piece of furniture and filling it in For example, the reflectivity of a conference table was measured to be 0.15, and the reflectivity of a metal bracket was measured to be 0.45.

[0079] parameter, Parameters and The steps to obtain the parameters are: Indicates the total number of reference positions when calculating lighting intensity. It depends on the number of measurement points arranged in the room. For example, in some scenes, additional measurement points will be arranged in the center, four corners and door of the room. If there are 6 measurement points in total, ; Indicates the position number to which the wall absorption coefficient is referenced. If a room has four walls and the paint or material of each wall is not exactly the same, it is necessary to test the four walls separately and record the absorption coefficient. ; Indicates the total number of furniture surfaces referenced. If there are 5 main pieces of furniture in the room and each is made of different materials, then .

[0080] Calculation process:

[0081] Suppose a room has 3 measuring points, denoted as , if there are two walls with the same material and the other two walls with different materials, a total of three absorption coefficients are recorded, corresponding to , if there are 3 pieces of furniture in the room , the activity intensity, natural light illumination and brightness values ​​of each measuring point are as follows: , , ; , , ; , , , the wall absorption coefficient is recorded as , , , the furniture reflection coefficient is recorded as , , , calculate the numerator first:

[0082]

[0083] Calculate the denominator again:

[0084]

[0085] Then calculate the furniture surface reflection coefficient term:

[0086]

[0087] Substituting the above results into the formula, we get:

[0088]

[0089] The result shows that a lighting intensity of about 4546.2246 lx is required. This value can be used to evaluate whether it is necessary to increase or decrease the brightness of the lamps or switch to different lamp combinations. If the subsequent calculated value is significantly greater than 6000 lx, it means that the room needs to be very bright. If it is less than 300 lx, it means that the room needs to be extremely low.

[0090] According to the lighting demand data, the required color temperature adjustment parameters are calculated using the following formula:

[0091]

[0092] in, Adjust the parameters for the desired color temperature, For the The current light source brightness at For the The current color temperature value at For the The window transmittance at For the The light source environment interference factor at is the preset target color temperature, is the color temperature measurement value of the current environment, The total number of light source points used as reference when calculating the color temperature adjustment parameters. Total number of references for windows and environmental distractions;

[0093] The benefit of the formula is that it incorporates the current light source brightness, window transmittance, and light source interference factor into the calculation of color temperature, and incorporates the deviation between the preset target color temperature and the actual measured value of the environment, so that the calculated color temperature adjustment parameters can both reflect the impact of external natural lighting and balance the difference between the preset color temperature and the actual color temperature.

[0094] The parameter acquisition step is the first The current light source brightness at the location is usually measured with a illuminance meter in the area where the LED lights are placed and converted into the corresponding light source brightness value. If there are multiple lamp heads, their brightness can be accumulated or position-weighted to obtain the comprehensive brightness. Pay more attention to the luminous ability of the light source itself. When the brightness of the LED lamp reaches 600lm and remains stable at a certain measuring point at 9 am, Input.

[0095] The parameter acquisition step is the first The current color temperature value at the lamp can be measured by placing a color thermometer near the lamp to record its color temperature. If a warm white LED lamp is installed, the color temperature can be detected to be around 3000K. If a cold white LED lamp is installed, the color temperature can be recorded to be 5000K or higher. By continuously recording the color temperature variation range of the lamp within a certain period of time, and then averaging the collected values, the color temperature in a stable state can be obtained. For example, after measuring the average color temperature at night for many times, it is found to be 4500K. .

[0096] The steps to obtain the parameters are as follows: The light transmittance of the windows at the building is 0.3. If the building windows have multiple glass materials, the light transmittance needs to be measured one by one. You can shine a fixed intensity of light on the outside of the window and measure the transmitted light intensity on the inside to get the light transmittance. If the light intensity on the outside is 10000lx and the measured light intensity on the inside is 3000lx, the light transmittance is 0.3. Record this data in If the room has three windows and their transmittances are measured to be 0.2, 0.25, and 0.3 respectively, they can be replaced in sequence in subsequent calculations.

[0097] The steps to obtain the parameter are as follows: The light source environment interference factor at the location is used to measure the degree of additional brightness interference caused by external ambient light, mutual interference or reflection of multiple light sources in the room. In actual measurement, the proportion of brightness contributed by another light source or natural light after a main light source is turned off or weakened can be recorded. If the proportion is high, it means that the interference factor is large. To obtain an example value, each lamp can be lit separately in a room and the impact of other lamps on the overall brightness can be measured, and then the situations of superimposing different lamps can be compared. If it is detected that a group of lamps interacts with natural light in a large range and causes the brightness of the measuring point to increase or decrease dramatically, the interference factor is set to 2 or 3. If the impact is stable, it is set to 1 or 0.5.

[0098] The parameters are obtained by setting the target color temperature in advance according to the room function. For example, the working environment is usually 4000K to 5000K, and the leisure environment is usually 2700K to 3500K. A relatively fixed value can be determined during the design stage or debugging. If the target color temperature of a conference room is set to 4500K, then .

[0099] The steps to obtain the parameter are as follows: This parameter represents the color temperature measurement value of the current environment. You can use the same color temperature meter to detect the center of the room. If the measured value is 4200K, then .

[0100] Parameters and The steps to obtain the parameters are: The total number of light sources used to calculate the color temperature adjustment parameters. For example, if there are 5 LED lights installed in the room and they are all in working condition, , is the total number of references for windows and environmental interference factors. If the room has 3 windows and 2 obvious external light source interference points, .

[0101] Calculation process:

[0102] Suppose there is a room The brightness of the main light source points is measured , and , The room has A window or environmental interference source, assuming that its transmittance and interference factor are recorded as , , , , and preset the target color temperature , current ambient color temperature , calculate the numerator first:

[0103]

[0104] Then calculate the interference term in the denominator:

[0105]

[0106] Substituting this into the denominator:

[0107]

[0108] Then calculate the color temperature difference:

[0109]

[0110] Substituting the above results into the formula:

[0111]

[0112] The result shows that a color temperature adjustment value of approximately 3655500.83 needs to be allocated to measure the difference between the color temperature of the room light source and the ambient color temperature. If the result obtained in the subsequent calculation exceeds 4 million, it means that the difference between the ambient and target color temperatures is large. If it is less than 100,000, it means that the two are close.

[0113] Based on the lighting intensity and color temperature adjustment parameters, the matching relationship between lighting requirements and activity types is analyzed, the data points with abnormal fluctuations in the calculated values ​​are corrected, the lighting intensity and color temperature adjustment parameters are determined, and the lighting adjustment requirements are generated.

[0114] Based on the lighting intensity and color temperature adjustment parameters obtained above, first match the lighting intensity value with the brightness level required by the activity type one by one, and check whether it is within the preset range. If a certain type of scene requires an ambient brightness of 100lx to 300lx, and the currently calculated lighting intensity is as high as 600lx, it means that there is an obvious deviation and the brightness value needs to be further adjusted. To determine the preset range, the ideal brightness requirements of various types of activities within seven days can be continuously measured and recorded before the room is delivered, and the highest and lowest values ​​can be compared and corrected. After comprehensive statistics, the target range of each activity type can be obtained. Then, the same comparison process is performed on the color temperature adjustment parameters, and the currently calculated color temperature adjustment amount is checked with the color temperature requirement threshold under different behavior scenarios. For example, for the work scene, the color temperature requirement can be set between 3500K and 4500K. By collecting the color temperature distribution of the room in a typical work scene, the acceptable minimum and maximum values ​​are screened out and set within the range of 3500K and 4500K. If the result of the current color temperature adjustment amount represents a large deviation, it is necessary to compare the specific color temperature of the lamp. The increase or decrease scheme, such as replacing the light source or controlling the output of the blue and red chips in the lamp, is then aggregated and analyzed for time segments with large numerical fluctuations. When analyzing the environmental interference data of these time periods, additional information such as the external window area and the light penetration of adjacent rooms can be collected to identify whether these interference factors cause abnormal fluctuations. If the color temperature is detected to change rapidly by more than 300K multiple times in one day, the time period and the corresponding light source position are recorded and included in the abnormal list. Then, by combining on-site measurements, it is determined whether it is caused by equipment failure or frequent user operation. The abnormal fluctuation point can be compared with the actual lighting state to confirm the numerical difference. If the difference is significant, the corresponding measurement parameters are re-corrected. If the difference is small, only normal records are made. Finally, a new light intensity and color temperature data pair is formed after eliminating significant anomalies. By comparing it with the interval of room requirements, check whether it meets the requirements of the corresponding activity type one by one. If the matching result shows that all data are within a reasonable range, it is determined that the lighting and color temperature requirements have been matched for the period. Finally, these corrected lighting intensity values ​​and color temperature adjustment parameters are combined to generate lighting adjustment requirements.

[0115] The steps to obtain the lamp adjustment results are:

[0116] Based on the lighting adjustment requirements, the target lighting intensity and target color temperature parameters in the lighting adjustment requirements are extracted, the brightness and color temperature setting values ​​of the current LED lamps are analyzed, and the difference between the two is calculated. It is analyzed whether it is necessary to adjust the brightness, color temperature separately or the brightness and color temperature at the same time, and the lighting adjustment instructions are obtained;

[0117] According to the lighting adjustment instruction, the control module of the LED lamp is called to send brightness and color temperature adjustment signals to the target lamp to form the lamp adjustment result.

[0118] Specifically, based on the target lighting intensity and target color temperature parameters included in the lighting adjustment requirements obtained earlier, first read and disassemble the numerical part and applicable time period information of these two data, and then record the brightness setting value and color temperature setting value of the current LED lamp in actual use. When comparing the brightness difference, a benchmark threshold can be set first. For example, in an office environment, the threshold is set to 100lx. This value comes from a survey of multiple users in the past week and combined with the brightness data statistics of each measuring point. If it is detected that the difference between the current brightness setting value and the target value exceeds 100lx, the position is deemed to not meet the preset requirements and is marked as requiring brightness adjustment. The same is true for the comparison of color temperature. The threshold can be temporarily set to 300K. This value is obtained by referring to the common perceptible color temperature variation range of the LED lamp and combining user feedback. When the comparison finds that the color temperature difference is greater than 300K, it is determined that color temperature adjustment needs to be performed. At the same time, it is necessary to analyze whether the previous lighting adjustment requirements require only brightness to be guaranteed or only adjustment to be prioritized. Color temperature: if the requirement document specifies that brightness should be met first, the brightness will be corrected first. If the priority is not specified in the requirement document, the deviation of brightness and color temperature will be compared. If the deviation of both exceeds the corresponding threshold, it is determined that the brightness and color temperature need to be adjusted at the same time. If only one item exceeds, adjust this item alone. After sorting out the target values ​​and current set values ​​of each time period and their differences, a set of instructions is formed according to the corresponding required adjustment form (individual brightness, individual color temperature or joint adjustment), and the expected values ​​of each parameter and the maximum allowable deviation range are indicated in the additional identification. For example, if 600lx brightness and 4000K color temperature are required for a certain time period, the brightness deviation range can be set to 50lx and the color temperature deviation range can be set to 200K. The specific values ​​are derived from the statistical analysis of the installation environment and the sensitivity threshold submitted by the user. Before executing the instruction, it is necessary to confirm the technical limitations of the lamp and record the excessive requirements that may not be met. Finally, the corresponding time period label and lamp number are added to each instruction to obtain the lighting adjustment instruction.

[0119] According to the lighting adjustment instructions obtained earlier, first extract the target lamp number pointed to by each instruction and its expected brightness and color temperature from the instruction set, and compare these expected values ​​with the adjustable range of the actual device one by one. For example, the brightness of LED lamps can be adjusted from 0lx to 2000lx, and the color temperature can be adjusted from 2700K to 6500K. If it is found that the brightness value or color temperature value set in the instruction exceeds the limit range of the lamp, it must be recorded before execution and the instruction must be marked as abnormal. If it is within the adjustable range, the corresponding value will be written into the sending queue, and then the instructions will be sent in sequence in the pre-set transmission channel in the form of digital signals to ensure the accuracy of the data. The data structure needs to be verified again before transmission, including all information such as time period identification, lamp number, brightness and color temperature target values, and threshold judgment conditions. If an incomplete signal or a disordered sequence is detected during the transmission process, the situation will be recorded and the transmission of this instruction will be suspended. After the inspection is completed, the brightness adjustment and color temperature adjustment instruction combinations will be sent in sequence. When the lamp receives it, a status readback operation is required. During the readback process, if it is found that the current lamp status is greater than the set threshold value The difference between the target value continues to send correction instructions until the measured difference falls back to within the threshold or reaches the limit range of the device. Finally, after all instructions are sent successfully, they are summarized to form the lamp adjustment result.

[0120] The steps to obtain the lighting mode optimization instructions are:

[0121] According to the lamp adjustment results, calculate the lighting mode deviation value, the calculation formula is:

[0122]

[0123] in, is the lighting mode deviation value, is the brightness value of the lamp in the current period, is the target brightness value, Adjust the impact factor for brightness, is the color temperature value of the current lamp, is the target color temperature value, is the ambient light source interference factor;

[0124] Based on the lighting mode deviation value, analyze the rationality of the lighting mode adjustment, determine whether the brightness and color temperature deviation of each lamp exceeds the set range, and generate lighting mode optimization instructions.

[0125] Specifically, the formula is beneficial in that, when evaluating the difference between the lighting mode and the target requirement, it takes into account both the brightness difference and the color temperature difference, and weights or corrects them through the brightness adjustment influencing factor and the ambient light source interference factor, so as to more accurately present the deviation between the current lighting state and the ideal lighting state.

[0126] The steps to obtain the parameter are as follows: This parameter represents the brightness value of the lamp in the current period. It can be collected by placing a light meter at a certain distance (for example, 1 meter) from the lamp. Each collection requires no additional changes to the environment around the lamp at the same time, so as to obtain a stable reading. If there are multiple lamps in the room, measure each one separately or focus on the main lamps. In order to quantify and incorporate the brightness value into this formula, it is necessary to first convert the light meter reading (unit lx) to the nominal output brightness level of the lamp itself. If the lamp has a direct brightness value display, it can be read directly. If you want to determine this parameter, an example is as follows: Place a light meter in the center of the room and record the current output brightness of the lamp as 750lx. This reading can be directly used as a value in some applications. It can also be converted according to the lumen output of the lamp itself. When multiple readings exist at the same time, the average value can be selected as the final .

[0127] The parameter acquisition steps are as follows: This parameter is the target brightness value, which usually comes from the target brightness information recorded in the lighting requirements or lamp adjustment results obtained previously. In order to assign a definite value, it is necessary to collect the range of users' or spaces' brightness requirements in the room usage scenario design stage, and set the typical value of this range as the standard brightness. For example, after a week of continuous collection in an office area, the average ideal brightness is about 500lx, so this value can be defined as the standard brightness during this period. In more specific scenarios, the target value can also be adjusted according to specific tasks. If a conference room requires 700lx during a certain period of time, .

[0128] The steps for obtaining the parameter are as follows: This parameter is the brightness adjustment influencing factor, which is used to reflect the different weights that may be brought to the overall lighting mode deviation under the same brightness difference conditions. Its value needs to be quantified after collecting data based on the room size, wall color, number of lamps, and distribution of lamps. The quantification method can refer to the following example: First test the impact of the illumination change at a specific location on the overall environment when multiple lamps in the room are turned on separately, and quantify the impact into a numerical value. If it is found that the room size is large and the influence range of a single lamp is limited, a smaller coefficient is assigned, such as about 0.5. If the space is small or the concentration of lamps is high, a larger coefficient can be assigned, such as 2. Take a specific calculation process as an example: In a 20-square-meter room, first record the center illumination difference when a single lamp is turned off and on, the change in wall light reflection, and the local brightness difference in the surrounding area. By accumulating and dividing by the reference area, an impact amplitude value of 1.2 is obtained, and then the mapping relationship for this value in historical experience is selected. .

[0129] The steps to obtain the parameter are as follows: This parameter refers to the color temperature value of the current lamp. The color temperature of the lamp light needs to be measured by a color thermometer at the same time point and its value (unit K) is recorded. To ensure the accuracy of the measurement, the color thermometer can be placed in an area that is a proper distance away from the lamp and is not interfered by other light sources. If a lamp in the room is a cool white lamp and the measured value is 5000K, then the value is directly recorded as .

[0130] The parameter acquisition step is: This parameter represents the target color temperature value, which is the same as the previously obtained Similarly, it is usually derived from the expected color temperature given in the lighting requirements or lamp adjustment results. If a rest area is determined to use a warm white temperature of 3500K to improve comfort after preliminary testing and user feedback, this value can be written into the system parameters to form .

[0131] The steps to obtain the parameter are as follows: This parameter is the ambient light source interference factor, which is mainly used to quantify the superposition or interference effect of external light in the room or from the window on the brightness and color temperature of the current lamp. It needs to be integrated with field monitoring and historical statistical processes. For example, if the natural light intensity in a room is observed to fluctuate greatly over time, and it is also observed that strong light sources in adjacent areas will produce obvious cross-light in the room, a higher interference factor needs to be assigned during the evaluation. The acquisition process can be as follows: first, regularly record the contribution of external light at multiple measuring points during the day, calculate the difference between indoor and outdoor lighting in the same period, and then compare the actual brightness generated by the lamp to see its proportion. If the external environment's light accounts for 50% of the total, the interference factor can be set to 1. If the external light is almost the same as the brightness of the lamp, it may be set to about 2. Finally, an example is used to illustrate that if the ambient light detected at 17:00 in the afternoon accounts for 30% of the total illumination, and the total interference ratio is measured at about 40% in combination with the indoor wall reflection, then the The agreed value is between 0.4 and 1.

[0132] Calculation process:

[0133] Set the current period of time lamp brightness value lx, target brightness value lx, brightness adjustment factor , current lamp color temperature value K, target color temperature K, ambient light source interference factor ;

[0134] First calculate the absolute value of the brightness difference and the square of the color temperature difference:

[0135]

[0136] Multiply it by the brightness adjustment factor:

[0137]

[0138] Substitute the above result into the numerator and calculate the denominator:

[0139]

[0140] Finally, we get:

[0141]

[0142] The results show that the lighting mode deviates significantly from the target in terms of brightness and color temperature, with the values ​​reaching over 900,000. Values ​​above a certain range (e.g. around 100,000) can indicate that a significant adjustment in brightness or color temperature is needed, while values ​​below several thousand indicate that the two are very close.

[0143] According to the lighting mode deviation value obtained above, the brightness collection records and color temperature collection records of all lamps in the current period are first disassembled and extracted, and the actual brightness and color temperature data corresponding to each lamp are compared one-to-one with the target value to check whether the deviation value obviously exceeds the reference range. For example, the preset deviation tolerance range can be compared. When the brightness deviation tolerance range is within 50lx and the current brightness difference of a certain lamp is as high as 120lx, the lamp is marked as needing further adjustment. The color temperature is also checked. An acceptable range of 300K can be defined. If the deviation measured at a certain point is 500K, it is considered to be seriously beyond expectations and the point is recorded. Subsequently, a more detailed investigation or retest is carried out on these marked lamps to rule out whether there is any measurement data in the timing or In case of mismatch in position, external light interference and room wall reflection should also be checked. After confirming that the measurement record matches the lamp number correctly, the ambient light interference information obtained earlier is introduced to analyze whether the outdoor sunlight intensity has changed significantly during this period. If it is found that it still deviates significantly from the target range after multiple time period comparisons, the lamp will be included in the list that needs to be tracked continuously. Combined with the lighting demand records obtained in the early stage, it is found whether it can be optimized by dimming in a small range or switching to a specific lamp combination. Records with obvious deviations in brightness and color temperature are centrally organized to form a mode adjustment comparison table to prompt subsequent operators or automatic programs to issue more stringent control instructions to these lamps. Finally, the system generates lighting mode optimization instructions one by one according to the classified deviations.

[0144] The steps to obtain environmental quality analysis results are:

[0145] Collect indoor and outdoor air quality and temperature and humidity data, integrate the data, and obtain pre-processed environmental data;

[0146] Based on the preprocessed environmental data, the comprehensive environmental quality index is calculated using the following formula:

[0147]

[0148] in, is the comprehensive index of environmental quality, is the indoor carbon dioxide concentration, is the outdoor ozone concentration, is the indoor humidity, is the comfortable humidity threshold for human body, is the indoor temperature, is the outdoor temperature, is the indoor PM2.5 concentration, is the outdoor PM10 concentration, is the wind speed;

[0149] Based on the comprehensive environmental quality index, the overall environmental quality status is evaluated, the periods of abnormal environmental quality are screened, and environmental quality analysis results are generated.

[0150] Specifically, based on the indoor and outdoor air quality and temperature and humidity data obtained previously, first record the indoor carbon dioxide concentration, indoor PM2.5 concentration, indoor humidity, indoor temperature and other items with the corresponding outdoor ozone concentration, outdoor PM10 concentration, outdoor temperature and wind speed in the same time series dimension, and keep each group of data consistent by matching the timestamp and geographic location code. Then, detect whether there are any abnormal phenomena in the values ​​that are missing or out of the sensor range. If a non-numeric value is detected or it significantly exceeds the set limit of the sensor, the abnormal point is recorded and replaced with the average of the previous three normal measurements. When using this processing method, statistics should be taken for the same time period in the past week, and a corresponding average or median should be calculated, and an increase or decrease should be given according to the fluctuation range detected by the actual sensor. The step value is used to complete the numerical correction. Subsequently, all aligned and anomaly-corrected data are normalized one by one for temperature, humidity, and pollutant concentration. For example, the temperature is compared with the range of 0℃ to 90℃. If a data record is 45℃, it is normalized to 0.5. If the humidity is compared with the range of 0% to 100% and recorded as 40%, it is normalized to 0.4. For pollutant concentration, the effective range of 0 micrograms per cubic meter to 500 micrograms per cubic meter can be referred to. The normalized value is calculated in the same way. Next, the integrated data set is arranged in a time series and marked for each time period at an hourly or finer frequency. When outdoor data is missing, the adjacent time period reference or multi-day comparison method is also used to fill it. After all collection items are normalized and time-series aligned, a complete set of pre-processed environmental data is obtained.

[0151] The benefit of the formula is that it simultaneously introduces multi-dimensional factors such as indoor and outdoor air pollution indicators, temperature and humidity differences, and wind speed, to comprehensively quantify environmental quality, which can provide a more intuitive numerical reference for subsequent screening of abnormal time periods and formulation of corresponding measures.

[0152] The steps to obtain the parameter are as follows: This parameter represents the indoor carbon dioxide concentration in ppm. It is necessary to place a sensor device that can record the carbon dioxide content in the room and maintain a certain sampling frequency. In order to obtain a reasonable range of values, it is necessary to first refer to the indoor crowd density. If the crowd is relatively concentrated, the measured value will increase. If the indoor ventilation is good, the value will decrease. In actual measurements, the sensor can be set to record the value every five minutes, and the measurement results within a day can be summarized after the sensor is calibrated. The extremely high or extremely low values ​​that appear in them are checked. After troubleshooting or abnormal conditions, the corresponding values ​​for each time period are selected. Used for difference calculation with ozone concentration, if the carbon dioxide concentration reaches 900ppm during the daytime peak period is detected in a certain occasion.

[0153] The steps for obtaining the parameter are as follows: This parameter refers to the outdoor ozone concentration, which can be expressed in ppb or μg / m3. It is necessary to set up a sampling point outside the building or check the real-time monitoring database of the local environmental protection department to obtain a reliable value. To maintain a consistent time period, the outdoor ozone concentration data and the indoor carbon dioxide data need to be recorded at the same time point, and then the indoor and outdoor difference is processed. Here, the external ozone concentration can be recorded by measuring instruments on the roof or in a designated area, or read using the data interface of an authoritative monitoring platform. If the monitoring shows that at 14:00 p.m. If the value is 60ppb, then this value is recorded in the corresponding time period.

[0154] The parameter acquisition steps are as follows: This parameter is the indoor humidity, the unit is usually a percentage, and the value is recorded in the specified space by an indoor hygrometer. To ensure the reliability of the data, it is necessary to configure the sensor in the main activity area of ​​the room and avoid directly facing the influence points of equipment such as air conditioner vents or humidifiers. The reasonable range of humidity is usually between 30% and 70%. The value is read every ten minutes and the time period mark is recorded for subsequent calculations. For example, at 10 am, the value is , measured at 14:00 p.m. .

[0155] The parameter acquisition steps are as follows: This parameter is the comfortable humidity threshold for humans, which is usually set at around 40% based on human body perception and health research, and can be fine-tuned within a certain range. The setting process can refer to relevant national or industry standard documents, and combine the experience of building air conditioning design departments to list the comfortable humidity range for humans in different seasons, and then take the middle value or common recommended value to define For example, actual measurements by relevant organizations show that most people feel relatively comfortable at 40% humidity, so Set to 40.

[0156] The steps for obtaining the parameter are as follows: the parameter is the indoor temperature, which is measured at key locations indoors using a thermometer or infrared thermometer. If the temperature distribution in a room is uneven, sensors can be placed in multiple locations and the measured values ​​can be weighted averaged. In practical applications, the central location or areas where people often move around can be selected as the main recording point. The temperature is in degrees Celsius, and the common reasonable range is between 18°C ​​and 30°C. The temperature value is read once in each time period and a judgment is made as to whether an extreme value occurs. For example, if 28°C is recorded in the afternoon in summer and 22°C is recorded in the morning, both can be written accordingly. .

[0157] The steps to obtain the parameter are as follows: This parameter represents the outdoor temperature, also in degrees Celsius, and can be recorded by placing a temperature sensor with a radiation shield in the shade outside the building. In order to reduce the additional interference caused by direct sunlight, it is necessary to choose a location that avoids direct sunlight and keep air circulation around the sensor. Repeat the measurement in multiple time periods to ensure that the data is consistent with the time node of the indoor temperature measurement, and finally bring the recorded value into the formula for absolute difference calculation.

[0158] The parameter acquisition steps are as follows: This parameter is the indoor PM2.5 concentration, which is usually expressed in μg / m3. Laser or infrared particle measuring instruments can be used to collect the data indoors and set at a height of about 1 meter from the ground. The sensor needs to be calibrated regularly to ensure data accuracy. The sampling frequency can be set to once every five minutes. Into the formula denominator, need to combine the same period and For example, a certain indoor environment shows It is 35μg / m3.

[0159] The parameter acquisition steps are as follows: This parameter represents the outdoor PM10 concentration, and the unit is also μg / m3. It is obtained by placing a measuring device outside the building or reading the public data of the local environmental monitoring department. In order to ensure that the values ​​correspond to the same time point, the outdoor PM10 value needs to be aligned with the indoor PM2.5 measurement time. If the outdoor PM10 concentration is monitored to be about 70μg / m3 at 15:00 on the same day, the value is written into the calculation of the corresponding time period.

[0160] The parameter acquisition steps are as follows: This parameter refers to wind speed, which can be monitored by an anemometer around the building or at a specific point outdoors, in m / s. To ensure data consistency, it is necessary to obtain the wind speed value at the same time after reading indoor PM2.5 and outdoor PM10 in the same measurement cycle. If the wind speed value changes between 0.5 and 2.5 m / s after three days of continuous monitoring, the corresponding value is placed in the time stamp of each record. For example, after multiple data are aligned at 14:00 in the afternoon, the wind speed is 1.5m / s. Substitute it into the operation of adding 1 to the denominator.

[0161] Calculation process:

[0162] First, from the pre-processed data obtained above, read the indoor carbon dioxide concentration at the same time , outdoor ozone concentration , Indoor humidity , Human Comfort Humidity Threshold , Indoor temperature , Outdoor temperature , Indoor PM2.5 concentration , Outdoor PM10 concentration and wind speed , respectively substitute the above values ​​into the numerator and denominator for calculation. The numerator contains the square root of the difference between carbon dioxide and ozone, the square term of indoor humidity and comfort threshold, and the absolute difference between indoor and outdoor temperature. The denominator is formed by adding PM2.5, PM10 and wind speed and then adding 1. Finally, divide the numerator and denominator to get the comprehensive environmental quality index. . Here is an example calculation:

[0163] make ppm, ppb, , , ℃, ℃, μg / m3, μg / m3, m / s, then calculate the numerator first:

[0164]

[0165]

[0166]

[0167] The sum of the molecules is ;

[0168] Next, calculate the denominator:

[0169]

[0170] Then divide the numerator by the denominator:

[0171]

[0172] The results show that at the current measurement time, the comprehensive environmental quality index is about 0.534. The acceptable range is set between 0.3 and 0.5. The current value slightly exceeds the upper limit, and a warning is required that the environment during this period may be less than ideal.

[0173] Based on the comprehensive environmental quality index obtained above, firstly, The values ​​are arranged in time series, and the index of each time period is compared with the pre-established judgment range. For example, 0.0 to 0.3 can be set as a good interval, 0.3 to 0.5 can be set as a general interval, and greater than 0.5 can be set as a poor interval. These thresholds are derived from the results of long-term statistics of the actual monitoring area and are corrected in combination with local indoor and outdoor air standards. If the index is higher than 0.5 in a certain period of time, it is marked as a potential environmental problem. If the value remains above 0.5, these periods are continuously recorded and the corresponding carbon dioxide concentration and PM2.5 and PM10 concentration details are extracted to check whether there are factors such as excessive equipment operation or insufficient ventilation. If the humidity difference or temperature difference measured by most sensors in the corresponding time period is also large, it means that there may be extreme environmental conditions, and manpower needs to be arranged for subsequent investigation. If the index is between 0.3 and 0.5, it means that the environment is relatively acceptable, and the value continues to be monitored to see if there is an upward trend. If there are multiple large jumps in one day, the short-term changes in wind speed and outdoor pollutant concentration are further compared, and all identified abnormal periods are centrally marked. Finally, these marks are summarized to obtain the environmental quality analysis results.

[0174] The steps to obtain ventilation adjustment results are:

[0175] Based on the environmental quality analysis results, by comparing with the predetermined health thresholds, the ventilation parameters that need to be adjusted are identified to form ventilation adjustment requirements;

[0176] Use the ventilation control interface to adjust ventilation volume and ventilation mode according to ventilation adjustment needs, including increasing or decreasing wind speed and switching from circulation mode to external ventilation mode;

[0177] Monitor the adjusted ventilation in real time, evaluate whether the effect of ventilation adjustment meets health and safety standards, and generate ventilation adjustment results.

[0178] Specifically, according to the environmental quality analysis results obtained above and the multiple reference items listed in the predetermined health thresholds, the data such as carbon dioxide concentration, humidity, temperature and PM2.5 are first disassembled and compared with the thresholds one by one. For example, if the acceptable range of carbon dioxide concentration is set between 300ppm and 1000ppm, each record is compared with the interval. If it is detected that the carbon dioxide concentration exceeds 1000ppm in a certain period of time, it is marked that the indoor air replacement frequency needs to be increased during this period. If the humidity comparison interval is set to 30% to 60% and a record is higher than 60% or lower than 30% in a certain period of time, the period is also marked. These marks are cross-reviewed with indoor and outdoor temperature difference, wind speed and other data to rule out abnormal deviations caused by measurement delays or sensor failures. Then all marked records are numbered one by one and combined with indoor and outdoor wind direction conditions to determine the specific ventilation parameter adjustment direction. For example, if the outdoor temperature is much higher than the indoor temperature and the humidity difference is large, the wind speed can be lowered in the setting table to avoid rapid heating of the indoor airflow. At the same time, if the carbon dioxide concentration is higher than the threshold, the ventilation volume still needs to be increased but a large temperature rise should be avoided as much as possible. Therefore, it is necessary to record a moderate range of ventilation rate values ​​in the parameter set. For example, by analyzing the relationship between ventilation volume and indoor temperature changes when the temperature is above 30°C in the past seven days, an incremental method of wind speed not exceeding 2m / s is selected. If the outdoor PM2.5 exceeds 75µg / m³ at certain times of the day, the external ventilation mode should not be turned on during the corresponding period, but internal circulation should be used to filter the air instead, and specific minute-level recommendations are given for the length of time to be filtered. All these ventilation parameters that need to be adjusted are finally summarized into a demand list, and listed one by one according to factors such as time period, wind speed adjustment range, mode switching conditions, etc., and finally confirmed by the system or management personnel to form a ventilation adjustment demand.

[0179] Using the determined ventilation adjustment requirements, the wind speed value corresponding to each time period is compared with the original ventilation mode. If it is found that the target wind speed is greater than the current set value, the incremental description is entered in the adjustment option. For example, 0.5m / s can be increased on the basis of the original 1m / s. If the target wind speed is lower than the current set value, the corresponding reduction amplitude is recorded and combined with the external air conditions to determine whether to continue to use the circulation mode or switch to external ventilation. When switching modes, it is necessary to first query the pre-calculated external air quality data and compare the current PM2.5, PM10 and temperature range. If the external environment data is within the acceptance range, it is switched to external ventilation. If the external air quality index exceeds the pre-set value, the external ventilation mode is switched to the external ventilation mode. If the threshold value is such as AQI100, the circulation mode is arranged in the demand, and then these adjustment instructions are submitted to the ventilation volume control device for execution in time sequence, and fine-tuned according to the recorded hourly crowd density. For example, when it is detected that the crowd is concentrated in a certain period of time and causes the indoor carbon dioxide concentration to exceed 1000ppm, the ventilation volume is increased to 2m³ / min. When the crowd density is less than 10 people, the ventilation volume is adjusted back to 1m³ / min. After that, it is covered one by one in the schedule and updated at any time according to the external meteorological changes, so that in each specific period of time, the wind speed can be increased or decreased or the circulation and external modes can be switched according to the instructions specified by the demand, and the actual ventilation volume and ventilation mode can be obtained.

[0180] Real-time monitoring of the ventilation system that has been adjusted is carried out. Monitoring equipment is distributed at the air supply outlet, return air outlet and the central position of the room to collect indicators such as temperature, humidity and carbon dioxide concentration. The data is read once every five minutes or so, and then these data are matched one by one with the target range recorded in the ventilation adjustment requirements. If the wind speed value is found to be lower than the planned wind speed range, a prompt will be given on the control interface and the node will be marked as an operation deviation. If the carbon dioxide concentration is continuously higher than the set threshold of 1000ppm, an additional mark will be added to the monitoring table and the operator will be required to observe whether the external ventilation mode is started normally. If the humidity fluctuation exceeds the set range of 30% to 60%, the accuracy of the ventilation volume adjustment is compared when the ventilation record period is reviewed to ensure that the possible causes of the deviation are identified. If it is detected that the outdoor air is relatively clean but the system has been in a circulating state, it is necessary to check the previous instructions to eliminate the situation where the instructions are not executed or the damper control fails. After multiple rounds of comparison, each record is sorted out according to the time series and all monitoring data and deviation information are summarized. Finally, by comparing the values ​​with the health and safety standards, the monitoring conclusion is generated on the control interface and the corresponding problem points are recorded to form the ventilation adjustment result.

[0181] The steps to obtain the ventilation optimization strategy are:

[0182] The ventilation adjustment results are input into the reinforcement learning model, and the optimized learning results are generated by comparing the adjustment results with the set environmental goals for learning and matching;

[0183] According to the optimized learning results, the ventilation settings are reconfigured to generate a ventilation optimization strategy.

[0184] Specifically, when the indoor humidity, temperature, carbon dioxide concentration, PM2.5 and other data corresponding to the ventilation adjustment results are input into the reinforcement learning model one by one, it is necessary to first break down the difference between the actual measured value and the predetermined environmental target in each time period, and digitize the difference of each indicator. For example, when the environmental target requires the carbon dioxide concentration to be lower than 1000ppm, the temperature to be in the range of 20℃ to 26℃, the humidity to be in the range of 40% to 60%, and the PM2.5 to be lower than 35µg / m³, the actual measured values ​​of the current time period can be compared one by one to see whether they exceed the threshold and how much they exceed, and these differences can be quantified into numerical values ​​to form reward or penalty signals. The model continuously adjusts internal parameters by iterating and learning for several rounds in the model. Each iteration records the corresponding relationship between the input difference information and the ventilation adjustment action output by the model. For example, if the ventilation action output by the model causes the measured data in the next period to deviate significantly from the target range, a low score is given in the scoring process. If the output action brings each indicator close to the target range, a high score is given. The model automatically corrects internal data based on these scores and continues to iterate. When the number of iterations reaches a pre-set upper limit, such as 1000 times, check whether the final action strategy satisfies the matching degree of each indicator within an acceptable range. If it continues to be higher than the set value, the model will automatically correct the internal data and continue to iterate. If the difference in thresholds continues, the iteration rounds will be increased or some parameters will be updated. After multiple tests, a relatively stable strategy output will be formed, and the output of each action and the corresponding difference results will be integrated to obtain the optimized learning results. The reinforcement learning model adopts a hierarchical reinforcement learning structure. The top layer is the training management layer, which is responsible for comparing the environmental data of each period with the preset goals during the training process to form reward or punishment signals. The middle layer is the strategy network layer, which includes an input layer, several fully connected hidden layers and an output layer. The input layer receives environmental status information (including indoor temperature, humidity, carbon dioxide concentration, PM2.5 value and external air Air quality related data), each hidden layer uses a certain ReLU activation function and is equipped with a fixed number of neurons, such as 128 neurons in the first hidden layer, 64 neurons in the second hidden layer, and 32 neurons in the third hidden layer. The output layer scores the executable ventilation actions (such as wind speed increase, wind speed reduction, switching circulation mode or external mode, etc.) and selects the action with the highest score as the current decision of the model; the network is connected to the training management layer through a feedback loop. The training management layer performs ventilation adjustment in the actual environment or simulated environment according to the output action instructions, and reads the temperature, humidity, PM2.5 concentration, carbon dioxide concentration and other data and compare the difference with the target interval, calculate the reward or penalty value and then pass it back to the strategy network layer for parameter update; the input data is spliced ​​in a fixed vector form, including the indoor and outdoor environmental indicators at the current moment and the adjustment action identifier at the previous moment, and the output data is a vector that scores all optional actions. Each action corresponds to a numerical score, and finally the deterministic strategy selects the action with the largest score; through multiple rounds of iterations, batch updates of weights and biases, the model can converge to a better strategy mapping under a specific threshold, and load this strategy network in subsequent implementation to automatically calculate the optimal ventilation decision at different time periods. .

[0185] According to the optimized learning results obtained above, the ventilation actions contained therein are first mapped to the actual executable range, including wind speed adjustment, temperature control and external ventilation period division. When disassembling the mapping data, it is necessary to first check the range of the environmental target value and check the action instructions under each period or trigger condition output by the model. If the model instruction in a certain period increases the wind speed to 2.5m / s but the maximum safety limit of the system for ventilation equipment is only 2m / s, a correction process is recorded to reduce the wind speed to 2m / s and keep the remaining actions unchanged. If the model instruction in a certain period switches to external ventilation but the PM2.5 concentration or temperature difference of the external air exceeds the preset threshold in the previous sampling analysis, for example, the upper limit of PM2.5 is set to 50µg / m³ and the actual measurement reaches 60µg / m³, then when reconfiguring the period table, the period is switched back to the circulation mode and its status is re-marked. When all periods have completed the action correction and matched the actual restrictions of the system equipment, these configurations are integrated into a ventilation optimization strategy and the wind speed setting, mode selection and corresponding reasons for each period are registered in sequence, and finally a configuration plan for subsequent automatic execution or manual review is formed.

[0186] The present invention provides an energy-saving and low-carbon control system, comprising:

[0187] Environmental monitoring module, which collects indoor and outdoor air quality, temperature and humidity data and generates environmental quality reports;

[0188] The lighting adjustment module calculates the required lighting intensity and color temperature adjustment parameters based on the natural light intensity and room activity requirements in the environmental quality report, and generates lighting adjustment requirements;

[0189] The ventilation control module adjusts the ventilation volume and ventilation mode of the ventilation equipment according to the environmental quality report and generates ventilation adjustment results;

[0190] The lamp adjustment module adjusts the brightness and color temperature of LED lamps based on lighting adjustment requirements, generates lamp adjustment results, compares the lamp adjustment results with the room usage pattern, identifies the deviation of the lighting pattern, and generates lighting pattern optimization instructions;

[0191] The optimization strategy module uses the ventilation adjustment results to perform reinforcement learning analysis, compares the deviation between the adjustment effect and the expected target, and generates a ventilation optimization strategy.

[0192] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. An artificial intelligence-based building energy-saving and low-carbon control method for operation period, characterized in that: The following steps are involved: Collect and analyze room usage and light intensity to obtain activity type recognition results; calculate required lighting intensity and color temperature adjustment parameters based on the natural light intensity and indoor activity requirements in the activity type recognition results, and generate lighting adjustment requirements; Based on the lighting adjustment requirements, the brightness and color temperature of the LED lamps are adjusted in real time to generate lamp adjustment results; the lamp adjustment results are compared with the room usage pattern to identify the deviation of the lighting pattern and generate lighting pattern optimization instructions; Monitor and analyze the environmental quality according to the indoor and outdoor air quality and temperature and humidity data to obtain the environmental quality analysis results; adjust the ventilation volume and ventilation mode of the ventilation equipment according to the environmental quality analysis results to generate ventilation adjustment results; Based on the ventilation adjustment results, reinforcement learning is used to analyze the deviation between the adjustment effect and the expected target, and a ventilation optimization strategy is generated.

2. The artificial intelligence-based building energy-saving and low-carbon control method for operation period according to claim 1 is characterized in that: The steps for obtaining the activity type recognition result are: Collect real-time light intensity and room usage data, including light intensity, room entry and exit frequency and residence time, to generate environmental monitoring data; Based on the environmental monitoring data, machine learning is used to learn and analyze the rules of activities in the room to obtain behavior pattern analysis results; According to the behavior pattern analysis result, the activity type in the room is identified and verified, and the activity type includes rest, work or meeting, and the activity type identification result is obtained.

3. The artificial intelligence-based building energy-saving and low-carbon control method for operation period according to claim 1 is characterized in that: The steps for obtaining the lighting adjustment requirement are as follows: Extract the natural light intensity information from the activity type recognition result, and analyze the light distribution of the room in different time periods in combination with the indoor activity needs, obtain the light change trend of each area and the indoor use scene characteristics, and obtain the lighting demand data; According to the lighting demand data, the required lighting intensity is calculated using the following formula: in, is the required lighting intensity, For the The intensity of indoor activities For the The natural light intensity at For the The current lighting brightness at is the average lighting brightness of the current environment, For the The wall light absorption coefficient at For the The light reflection coefficient of the furniture surface at The total number of positions used for calculating lighting intensity. is the position number to which the wall absorption coefficient is referred, The total number of furniture surfaces referenced; According to the lighting demand data, the required color temperature adjustment parameters are calculated, and the calculation formula is: in, Adjust the parameters for the desired color temperature, For the The current light source brightness at For the The current color temperature value at For the The window transmittance at For the The light source environment interference factor at is the preset target color temperature, is the color temperature measurement value of the current environment, The total number of light source points used as reference when calculating the color temperature adjustment parameters. Total number of references for windows and environmental distractions; Based on the lighting intensity and color temperature adjustment parameters, the matching relationship between the lighting demand and the activity type is analyzed, the data points with abnormal fluctuations in the calculated values ​​are corrected, the lighting intensity and color temperature adjustment parameters are determined, and the lighting adjustment demand is generated.

4. The method for controlling energy-saving and low-carbon operation-period buildings based on artificial intelligence according to claim 1 is characterized in that: The steps for obtaining the lamp adjustment result are: Based on the lighting adjustment requirement, extract the target lighting intensity and target color temperature parameters in the lighting adjustment requirement, analyze the brightness and color temperature setting values ​​of the current LED lamp, calculate the difference between the two, analyze whether it is necessary to adjust the brightness, color temperature separately or the brightness and color temperature at the same time, and obtain the lighting adjustment instruction; According to the lighting adjustment instruction, the control module of the LED lamp is called to send brightness and color temperature adjustment signals to the target lamp to form a lamp adjustment result.

5. The method for controlling energy-saving and low-carbon operation-period buildings based on artificial intelligence according to claim 1 is characterized in that: The steps of obtaining the lighting mode optimization instruction are as follows: According to the lamp adjustment result, the lighting mode deviation value is calculated, and the calculation formula is: in, is the lighting mode deviation value, is the brightness value of the lamp in the current period, is the target brightness value, Adjust the impact factor for brightness, is the color temperature value of the current lamp, is the target color temperature value, is the ambient light source interference factor; Based on the lighting mode deviation value, the rationality of the lighting mode adjustment is analyzed, and it is determined whether the brightness and color temperature deviations of each lamp exceed the set range, and a lighting mode optimization instruction is generated.

6. The artificial intelligence-based building energy-saving and low-carbon control method for operation period according to claim 1 is characterized in that: The steps for obtaining the environmental quality analysis results are: Collect indoor and outdoor air quality and temperature and humidity data, integrate the data, and obtain pre-processed environmental data; Based on the pre-processed environmental data, the comprehensive environmental quality index is calculated using the following formula: in, is the comprehensive index of environmental quality, is the indoor carbon dioxide concentration, is the outdoor ozone concentration, is the indoor humidity, is the comfortable humidity threshold for human body, is the indoor temperature, is the outdoor temperature, is the indoor PM2.5 concentration, is the outdoor PM10 concentration, is the wind speed; Based on the comprehensive environmental quality index, the state of the overall environmental quality is evaluated, time periods with abnormal environmental quality are screened, and environmental quality analysis results are generated.

7. The artificial intelligence-based building energy-saving and low-carbon control method for operation period according to claim 1 is characterized in that: The steps for obtaining the ventilation adjustment result are: According to the environmental quality analysis result, by comparing with the predetermined health threshold, the ventilation parameters that need to be adjusted are identified to form ventilation adjustment requirements; Using the ventilation control interface, adjusting the ventilation volume and ventilation mode according to the ventilation adjustment requirements, including increasing or decreasing the wind speed and switching from the circulation mode to the external ventilation mode; Monitor the adjusted ventilation in real time, evaluate whether the effect of ventilation adjustment meets health and safety standards, and generate ventilation adjustment results.

8. The artificial intelligence-based building energy-saving and low-carbon control method for operation period according to claim 1 is characterized in that: The steps for obtaining the ventilation optimization strategy are: Inputting the ventilation adjustment result into the reinforcement learning model, learning and matching the adjustment result with the set environmental target by comparing the adjustment result, and generating an optimized learning result; According to the optimized learning results, the ventilation settings are reconfigured to generate a ventilation optimization strategy.

9. The energy-saving and low-carbon control system of the operation-period building energy-saving and low-carbon control method based on artificial intelligence according to any one of claims 1 to 8 is characterized in that: include: Environmental monitoring module, which collects indoor and outdoor air quality, temperature and humidity data and generates environmental quality reports; The lighting adjustment module calculates the required lighting intensity and color temperature adjustment parameters based on the natural light intensity and room activity requirements in the environmental quality report, and generates lighting adjustment requirements; The ventilation control module adjusts the ventilation volume and ventilation mode of the ventilation equipment according to the environmental quality report and generates ventilation adjustment results; The lamp adjustment module adjusts the brightness and color temperature of LED lamps based on lighting adjustment requirements, generates lamp adjustment results, compares the lamp adjustment results with the room usage pattern, identifies the deviation of the lighting pattern, and generates lighting pattern optimization instructions; The optimization strategy module uses the ventilation adjustment results to perform reinforcement learning analysis, compares the deviation between the adjustment effect and the expected target, and generates a ventilation optimization strategy.

Images