AI-based energy-saving lamp control method

Abstract

This invention discloses an AI-based energy-saving lamp control method, relating to the field of energy-saving lamp control technology, comprising the following steps: deploying various sensors in the classroom to collect data on the presence and behavior of people within the illumination range of the energy-saving lamps, and collecting light intensity control data of the energy-saving lamps; constructing a prediction model based on the presence data and performing presence prediction analysis to obtain presence prediction data; performing light intensity analysis based on the behavior data and light intensity control data to obtain optimal light intensity data; and controlling and adjusting the energy-saving lamps according to the presence prediction data and the optimal light intensity data. This invention addresses the problem that existing energy-saving lamp control technologies cannot predict the presence of people in a study classroom based on historical data, and cannot control and adjust the switching on / off of lamps and light intensity in the activity area based on the activity status of people.

Description

Technical Field

[0001] This invention relates to the field of energy-saving lamp control technology, specifically to an AI-based energy-saving lamp control method. Background Technology

[0002] Energy-saving lamp control technology refers to a comprehensive technical system that uses electronic circuit design, intelligent algorithms, and sensor integration to precisely regulate the working status of energy-saving lamps in order to achieve energy saving, efficiency improvement, extended lifespan, and enhanced ease of use.

[0003] Existing energy-saving lamp control technologies in study classrooms often rely on human presence sensors to determine the presence of people and adjust the lamps accordingly. This involves predicting the number of people in the classroom and adjusting the lamp brightness based on their presence. Furthermore, control based on human presence sensors often involves switching all the lamps on and off, failing to dynamically adjust the lighting area based on people's positions and activity areas, or automatically adjust the optimal light intensity based on user behavior and needs. For example, patent application CN116193682A discloses a classroom lighting control method and device that controls the lamps by determining the presence of people, but it cannot predict the number of people in the classroom or dynamically adjust the lighting area and intensity based on their positions and activity areas. Therefore, existing energy-saving lamp control technologies cannot predict the number of people in the classroom based on historical data, nor can they adjust the lamp switching and light intensity in the activity areas based on people's activity status. Summary of the Invention

[0004] This invention aims to at least partially solve one of the technical problems in the prior art. It involves deploying multiple sensors in the classroom to collect data on the presence and behavior of people within the illumination range of energy-saving lamps, as well as data on the light intensity control of the lamps. A predictive model is constructed, and a prediction analysis of the presence of people is performed to obtain predicted presence data. Furthermore, light intensity analysis is conducted to obtain optimal light intensity data. Based on the predicted presence data and the optimal light intensity data, the energy-saving lamps are controlled and adjusted. This addresses the problem that existing energy-saving lamp control technologies cannot predict the presence of people in the classroom based on historical data, nor can they control and adjust the switching on and off of lamps and the light intensity in the activity area based on the activity status of the people.

[0005] To achieve the above objectives, this application provides an AI-based energy-saving lamp control method, comprising the following steps:

[0006] Multiple sensors are installed in the classroom to collect data on the presence and behavior of people within the illumination range of the energy-saving lamps, as well as data on the light intensity control of the energy-saving lamps.

[0007] A predictive model is built based on personnel presence data, and personnel presence prediction analysis is performed to obtain personnel presence prediction data;

[0008] Based on human behavior data and light intensity control data, light intensity analysis was performed to obtain the optimal light intensity data;

[0009] The energy-saving lamps are controlled and adjusted based on the predicted personnel presence data and the optimal light intensity data.

[0010] Furthermore, various sensors are deployed in the classroom to collect data on the presence and behavior of people within the illumination range of the energy-saving lamps, and to collect data on the light intensity control of the energy-saving lamps, including the following sub-steps:

[0011] Human presence sensors, light intensity sensors, and image sensors are deployed in the classroom; any energy-saving lamp in the classroom is designated as the first energy-saving lamp, and the illumination range of the first energy-saving lamp is designated as the first illumination range.

[0012] Obtain the theoretical range of indoor light intensity and divide it into k1 indoor light intensity levels, where k1 is the number of levels set;

[0013] A human presence sensor is used to collect data on the presence of people within a first lighting range at a first time interval, and the time of collection is recorded. This data is denoted as the human presence data of the first energy-saving lamp, where the first time interval is t1.

[0014] Furthermore, the process of deploying various sensors in the classroom to collect data on the presence and behavior of people within the illumination range of the energy-saving lamps, as well as data on the light intensity control of the energy-saving lamps, also includes the following sub-steps:

[0015] If there are people within the first lighting range, the image sensor is used to collect behavioral images of each person within the first lighting range at a first time interval, and synchronously with the human presence sensor. The behavioral image of each person is analyzed to obtain the behavioral type of each person, which is recorded as the human behavior data of the first energy-saving lamp. The behavioral types of the people include learning behavior, resting behavior and interactive behavior.

[0016] If there are people within the first lighting range, the illuminance level of the first energy-saving lamp and the indoor light intensity level are collected synchronously with the human presence sensor at a first time interval and recorded as light intensity control data.

[0017] Furthermore, based on the personnel presence data, a predictive model is constructed, and personnel presence predictive analysis is performed to obtain personnel presence predictive data, including the following sub-steps:

[0018] The presence of personnel within the first lighting range in the personnel presence data is recorded as 1, and the absence of personnel is recorded as 0. The data is then sorted according to the corresponding date and collection time, and divided according to the corresponding date to obtain the personnel presence time series data.

[0019] For any given date, the time sequence data of personnel is arranged in ascending order of time and denoted as the first personnel existence sequence; the second time period is set to ZT2.

[0020] The sequence of the first person is divided into multiple subsequences with a time length of ZT2 using ZT2. Any subsequence is denoted as the first subsequence, and the time period corresponding to the first subsequence is denoted as the first time period.

[0021] Furthermore, the process of constructing a predictive model based on personnel presence data and performing personnel presence predictive analysis to obtain personnel presence predictive data also includes the following sub-steps:

[0022] For the first subsequence, obtain the proportion coefficient of 1 in the first subsequence, and denot it as the personnel coefficient of the first subsequence; repeatedly obtain the personnel coefficients of all subsequences in the first personnel existence sequence, and arrange them according to the corresponding time order, and denot them as the personnel coefficient sequence of the first personnel existence sequence;

[0023] Repeatedly obtain the personnel coefficient sequence for all dates in the personnel data, and concatenate them sequentially according to date order, denoted as the first personnel coefficient data;

[0024] For the first subsequence, obtain the subsequence corresponding to the first time period in other dates, arrange them in date order, and record them as the personnel coefficient data of the first time period. Repeat this process to obtain the personnel coefficient data of all time periods and record them as the second personnel coefficient data.

[0025] Furthermore, the process of constructing a predictive model based on personnel presence data and performing personnel presence predictive analysis to obtain personnel presence predictive data also includes the following sub-steps:

[0026] The Long Short-Term Memory Network was trained using the first and second personnel coefficient data, respectively, and the first and second prediction models were obtained after completion.

[0027] The mean absolute errors of the first prediction model and the second prediction model are obtained respectively and denoted as MA1 and MA2 in order.

[0028] Using the first personnel coefficient data and the first prediction model, the personnel coefficient data for all time periods over the next k2 dates are predicted to obtain the first prediction data, where k2 is the set number of dates;

[0029] Predict the personnel coefficient data for each time period in the next k2 dates using the second personnel coefficient data and the second prediction model, and arrange the personnel coefficient data for each time period predicted for each future date in the corresponding chronological order to obtain the personnel coefficient data for each future date, denoted as the second prediction data;

[0030] For any time period in the next k2 dates, denoted as the second time period; respectively obtain the corresponding data of the first prediction data and the second prediction data in the second time period, and denote them as AE1 and AE2 in sequence;

[0031] Calculate AE0, denoted as the final personnel coefficient of the second time period, where AE0 = Q1 * AE1 + Q2 * AE2, Q1 = MA2 / (MA1 + MA2), Q2 = MA1 / (MA1 + MA2); and calculate MA0, MA0 = Q1 * MA1 + Q2 * MA2;

[0032] For the second time period, if AE0 < MA0, mark the second time period as an unoccupied period, otherwise mark the second time period as an occupied period. Repeat the marking for all time periods in the next k2 dates. After completion, obtain the personnel presence prediction data.

[0033] Furthermore, based on the personnel behavior data and the light intensity control data, conduct light intensity analysis to obtain the optimal light intensity data, including the following sub-steps:

[0034] For any acquisition moment in the personnel behavior data, denoted as the first moment, obtain the total number of personnel collected at the first moment, and calculate the personnel ratios of learning behavior, rest behavior, and interaction behavior at the first moment respectively, denoted as PA1, PA2, and PA3 in sequence;

[0035] If PA1 = PA2 = PA3, mark the first moment as a learning behavior; otherwise, select the two largest among PA1, PA2, and PA3, and denote them as VA1 and VA2 in descending order;

[0036] In the case where VA1 is equal to VA2, if VA1 and VA2 include PA1, mark the first moment as a learning behavior, otherwise the first moment is an interaction behavior.

[0037] Furthermore, based on the personnel behavior data and the light intensity control data, conduct light intensity analysis to obtain the optimal light intensity data, which also includes the following sub-steps:

[0038] If VA1 is not equal to VA2, and VA1 and VA2 include PA1 and PA1 > k3*(VA2+VA1), then the first time step is marked as the learning behavior; otherwise, the first time step is marked as the behavior type corresponding to max(VA1, VA2), where k3 is the set scaling factor, k3≤0.5;

[0039] Repeatedly mark all collection times in the personnel behavior data to obtain the first behavior data;

[0040] The learning behavior, rest behavior, and interaction behavior in the first set of data are recorded as 00, 01, and 10 respectively, and sorted according to the corresponding date and collection time to form the time series data of personnel behavior.

[0041] Furthermore, the analysis of light intensity based on personnel behavior data and light intensity control data to obtain optimal light intensity data also includes the following sub-steps:

[0042] For any two adjacent data points in the time series data of personnel behavior on any given date, if the two data points are different, the middle of the two data points is used as the dividing position. All dividing positions are obtained repeatedly and then divided to obtain the first row sequence.

[0043] For any two adjacent division positions in the first row sequence, the corresponding behavior time period is marked according to the data contained in the two division positions. The behavior time period includes learning time period, rest time period and interaction time period.

[0044] For any given time period, denoted as the first time period, the illuminance level of the first energy-saving lamp corresponding to each collection moment in the first time period is obtained and denoted as the first light intensity sequence.

[0045] Obtain the proportion of each light intensity level in the first light intensity sequence, and obtain the light intensity level with the largest proportion, which is recorded as the first light intensity level. At the same time, obtain the indoor light intensity level collected at the sampling time corresponding to each light intensity level with the largest proportion, and calculate the average value, which is recorded as the indoor reference light intensity. Record the first light intensity level as the best light intensity under the indoor reference light intensity for the first row of time periods.

[0046] Repeatedly acquire the indoor reference light intensity and optimal light intensity corresponding to all time periods in the first row sequence, and record them as the optimal light intensity data for the corresponding date of the first row sequence; and repeatedly acquire the optimal light intensity data for all dates.

[0047] Furthermore, the control and adjustment of energy-saving lamps based on predicted personnel presence data and optimal light intensity data includes the following sub-steps:

[0048] For the current moment, the time period corresponding to the personnel presence prediction data is recorded as the current time period; if there are no personnel within the first lighting range in the current time period, and the next time period adjacent to the current time period in the personnel presence prediction data is an unoccupied time period, then the first energy-saving lamp is turned off.

[0049] If there are people within the first lighting range during the current time period, or if the next time period adjacent to the current time period in the prediction data is a time period with people, then turn on the first energy-saving lamp;

[0050] At the current moment, with the first energy-saving lamp on, the indoor light intensity is obtained and recorded as the current indoor light intensity HD; if there are people within the first lighting range, the corresponding behavior type at the current moment is obtained and recorded as the current behavior; if there are no people within the first lighting range, the learned behavior is recorded as the current behavior.

[0051] Get the best light intensity data within k4 days before the current time, and denot it as the reference best light intensity data; from the reference best light intensity data, get the two best light intensities with the smallest difference from the current indoor light intensity during the time period corresponding to the current behavior, and denot them as LA and LB respectively; and denot the indoor reference light intensities corresponding to LA and LB as HA and HB respectively in order, where k4 is the number of settings.

[0052] Calculate LZ and adjust the illuminance of the first energy-saving lamp to LZ, where LZ = Q3 * LA + Q4 * LB, where if HB - HD = HA - HD = 0, then Q3 = Q4 = 0.5, otherwise Q3 = |HB - HD| / (|HB - HD| + |HA - HD|), Q4 = 1 - Q3;

[0053] Repeatedly control all energy-saving lights in the classroom.

[0054] The beneficial effects of this invention are as follows: This invention collects data on the presence and behavior of people within the illumination range of energy-saving lamps by arranging various sensors in the classroom, and also collects light intensity control data of the energy-saving lamps; a prediction model is constructed based on the presence data, and presence prediction analysis is performed to obtain presence prediction data; light intensity analysis is performed based on the behavior data and light intensity control data to obtain optimal light intensity data; and the energy-saving lamps are controlled and adjusted according to the presence prediction data and optimal light intensity data. When controlling the energy-saving lamps in the self-study classroom, the situation of people in the classroom can be predicted based on historical data, and the switching on and off of lamps and the light intensity in the activity area can be controlled and adjusted according to the activity status of people.

[0055] This invention constructs a first personnel coefficient data set by sequentially concatenating daily personnel coefficients, and a second personnel coefficient data set by arranging personnel coefficients at the same time each day. Two models are trained separately to capture the overall dynamic changes in personnel flow throughout the day, as well as the periodic characteristics of individual moments. The results are weighted and averaged based on the mean absolute error, improving the accuracy and robustness of the predictions. By considering personnel conditions, behavior, and indoor / outdoor light intensity, different optimal lighting conditions are provided for different areas and behaviors of people, achieving precise control of the lighting system, improving energy efficiency, reducing energy consumption, and meeting users' personalized lighting environment needs. Attached Figure Description

[0056] Figure 1 This is a flowchart of the steps of the method of the present invention;

[0057] Figure 2 This is a flowchart of the first time-point marking process of the present invention;

[0058] Figure 3 This is a flowchart of the energy-saving lamp control and adjustment process of the present invention;

[0059] Figure 4 This is a schematic diagram of the electronic device of the present invention. Detailed Implementation

[0060] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0061] Example 1, please refer to Figure 1 As shown, this application provides an AI-based energy-saving lamp control method, including the following steps:

[0062] Step S1 involves deploying various sensors in the classroom to collect data on the presence and behavior of people within the illumination range of the energy-saving lamps, as well as data on the light intensity control of the energy-saving lamps. Step S1 includes the following sub-steps:

[0063] Step S101: Deploy a human presence sensor, a light intensity sensor, and an image sensor in the classroom; designate any energy-saving lamp in the classroom as the first energy-saving lamp, and designate the illumination range of the first energy-saving lamp as the first illumination range; the human presence sensor is used to detect the presence and activity of people in the space, the image sensor is used to analyze the environmental scene and personnel behavior; the light intensity sensor is used to record the current output intensity of the lamp and the ambient light level, providing data for subsequent lighting adjustments;

[0064] Step S102: Obtain the theoretical range of indoor light intensity and divide it into k1 indoor light intensity levels, where k1 is the number of levels set; Discretize the continuous light intensity values ​​to facilitate subsequent data comparison and fusion. k1 can be set according to the actual application scenario. The larger k1 is, the more refined the subsequent adjustments will be, but the computing power required will also be greater.

[0065] Step S103: Use a human presence sensor to collect data on whether there are people within the first lighting range at a first time interval, and record the collection time as the human presence data of the first energy-saving lamp, where the first time interval is t1; in this embodiment, t1 is 1 minute.

[0066] Step S104: If there are people within the first lighting range, use an image sensor to collect behavioral images of each person within the first lighting range at a first time interval, synchronously with the human presence sensor. Analyze the behavioral images of each person to obtain their behavioral type, which is recorded as the personnel behavior data of the first energy-saving lamp. The personnel behavior types include learning behavior, resting behavior, and interactive behavior. Learning behavior refers to active or passive learning activities centered around knowledge acquisition and skill training, and is the main personnel behavior in the classroom. Resting behavior refers to the behavior of people temporarily leaving the learning state to relax or adjust their personal state, usually a short, low-intensity activity, mainly including closing their eyes to rest and resting briefly on the table. Interactive behavior refers to other behaviors besides learning behavior and resting behavior, which are between learning and resting, mainly interpersonal or environmental interaction.

[0067] Step S105: If there are people in the first lighting range, use the light intensity sensor to collect the light intensity level of the first energy-saving lamp and the indoor light intensity level at a first time interval, and synchronously with the human presence sensor, and record them as light intensity control data.

[0068] In the specific implementation process, the illumination requirements vary under different activity states. For example, higher illumination is needed when studying, while it can be appropriately reduced when resting or discussing. The behavior classification results will serve as an important reference feature for subsequent processing, and dynamic lighting will be adjusted based on the behavior of people within the range. Furthermore, the classification of people's behavior can also be customized according to the actual application scenario.

[0069] Step S2 involves constructing a prediction model based on the personnel presence data and performing personnel presence prediction analysis to obtain personnel presence prediction data. Step S2 includes the following sub-steps:

[0070] Step S201: Record the presence of personnel within the first lighting range in the personnel presence data as 1 and the absence of personnel as 0, and sort them according to the corresponding date and collection time, and divide them according to the corresponding date to obtain the personnel presence time series data; through unified 0 / 1 encoding and precise time labels, a clear time series structure is provided for subsequent model input;

[0071] Step S202: For any date, the time sequence data of personnel existence is arranged in order from oldest to newest time and recorded as the first personnel existence sequence; the second time period is set to ZT2; in this embodiment, the second time period is 20 minutes.

[0072] Step S203: Divide the first personnel existence sequence into multiple subsequences with a time length of ZT2 using ZT2. For any subsequence, it is denoted as the first subsequence, and the time period corresponding to the first subsequence is denoted as the first time period.

[0073] Step S204: For the first subsequence, obtain the proportion coefficient of 1 in the first subsequence, and record it as the personnel coefficient of the first subsequence; repeatedly obtain the personnel coefficients of all subsequences in the first personnel existence sequence, and arrange them according to the corresponding time order, and record them as the personnel coefficient sequence of the first personnel existence sequence; compress the discrete 0 / 1 sequence into a continuous proportion index, which not only reduces the sparsity of the model input, but also intuitively reflects the characteristics of personnel density change in each time period.

[0074] Step S205: Repeatedly obtain the personnel coefficient sequence for all dates in the personnel existence data, and connect them sequentially according to date order, denoted as the first personnel coefficient data; construct a training set with a long time span so that the subsequent model can learn the macro-change pattern of personnel coefficient with date; such as the overall attendance changes at the beginning and end of the semester, before and after holidays, and different days of the week; classroom personnel attendance often shows a slow evolution, and long sequences can allow the model to learn these slow trends, thereby better predicting the overall attendance level for the next few days.

[0075] Step S206: For the first subsequence, obtain the subsequence corresponding to the first time period in other dates, arrange them in date order, and record them as the personnel coefficient data of the first time period. Repeat the process of obtaining the personnel coefficient data of all time periods and record them as the second personnel coefficient data. Construct an independent time series input for each specific time period, for example, 8:00–8:20, to capture the regular fluctuations of this time period in different dates. The attendance of people in the classroom often shows a highly similar attendance pattern in the same time period, such as before the start of class every morning. After aggregation and arrangement, it helps to accurately capture this periodic fluctuation. Compared with long trend data spliced ​​across dates, single time period series have lower noise, and the model input is more focused on the "fluctuation" characteristics of this time period, improving the accuracy of short-term prediction.

[0076] Step S207: Use the first personnel coefficient data and the second personnel coefficient data to train the long short-term memory network model respectively. After completion, obtain the first prediction model and the second prediction model respectively. That is, the first prediction model is trained by the first personnel coefficient data to capture the long-term trend. That is, the second prediction model is trained by the second personnel coefficient data to capture the change law within the time period. The dual-model architecture takes into account both the global trend and local fluctuations, and can not only learn the long-term trend but also capture the fine-grained changes in a specific time period.

[0077] Step S208: Obtain the mean absolute error of the first prediction model and the second prediction model respectively, and record them as MA1 and MA2 in sequence; quantify the prediction accuracy of the model to provide a basis for subsequent weighted fusion; the smaller the error of the model, the higher its prediction credibility.

[0078] Step S209: Use the first personnel coefficient data and the first prediction model to predict the personnel coefficient data for all time periods of the next k2 dates, and obtain the first prediction data, where k2 is the set number of dates; k2 can be set according to the actual application scenario. In this embodiment, k2 = 3, that is, predict the data for the next 3 days.

[0079] Step S210: Use the second personnel coefficient data and the second prediction model to predict the personnel coefficient data for each time period in the next k2 dates, and arrange the personnel coefficient data for each time period predicted for each future date in the corresponding time sequence to obtain the personnel coefficient data for each future date, recorded as the second prediction data; when the second prediction model makes a prediction, it needs to predict each future time period separately and then stitch them together to obtain continuous data.

[0080] Step S211: For any time period in the next k2 dates, denoted as the second time period; obtain the corresponding data of the first prediction data and the second prediction data in the second time period respectively, and record them as AE1 and AE2 in sequence.

[0081] Step S212: Calculate AE0, denoted as the final personnel coefficient for the second time period, where AE0 = Q1 * AE1 + Q2 * AE2, Q1 = MA2 / (MA1 + MA2), Q2 = MA / (MA1 + MA2); and calculate MA0, MA0 = Q1 * MA1 + Q2 * MA2; automatically assign weights according to the historical errors of each model to make the fusion prediction have both short-term accuracy and long-term stability.

[0082] Step S213: For the second time period, if AE0 < MA0, mark the second time period as an unoccupied period, otherwise mark the second time period as an occupied period. Repeat marking all time periods in the next k2 dates to obtain the personnel presence prediction data.

[0083] In the specific implementation process, the first personnel coefficient data and the second personnel coefficient data can be divided into training and testing parts, respectively, and used for training and testing to obtain the corresponding mean absolute error. The MA0 is used as the judgment standard to control the uncertainty of the prediction results and reduce the risk of misjudgment caused by model overconfidence. This provides a reliable judgment for the subsequent decision to switch energy-saving lamps on and off.

[0084] Step S3 involves analyzing the light intensity based on personnel behavior data and light intensity control data to obtain the optimal light intensity data. Step S3 includes the following sub-steps:

[0085] For step S301, please refer to... Figure 2 As shown, for any collection time in the personnel behavior data, it is recorded as the first time. The total number of personnel collected at the first time is obtained, and the proportion of personnel with learning behavior, rest behavior and interactive behavior at the first time is calculated respectively and recorded as PA1, PA2 and PA3 in order.

[0086] Step S302: If PA1=PA2=PA3, then mark the first moment as the learning behavior; otherwise, select the two largest values ​​among PA1, PA2 and PA3, and record them as VA1 and VA2 respectively in descending order; to ensure that learning behavior is given priority when the number of participants is equal.

[0087] Step S303: When VA1 equals VA2, if PA1 is included in both VA1 and VA2, then the first moment is marked as a learning behavior; otherwise, the first moment is marked as an interactive behavior. Interactive behaviors require more lighting than resting behaviors.

[0088] Step S304: If VA1 is not equal to VA2, and VA1 and VA2 include PA1 and PA1 > k3*(VA2+VA1), then the first moment is marked as a learning behavior; otherwise, the first moment is marked as the behavior type corresponding to max(VA1, VA2). For example, if max(VA1, VA2) = PA2, then the first moment is marked as a resting behavior. Here, k3 is a set proportional coefficient, k3≤0.5; in this embodiment, k=0.4.

[0089] Since classrooms are the primary learning spaces, and learning activities require more lighting than resting or interactive activities, a significant portion of people are in a learning state. Even if they don't constitute the majority, we can still determine that the first moment is a learning activity, and therefore increase the lighting accordingly.

[0090] Step S305: Repeatedly mark all collection times in the personnel behavior data to obtain the first behavior data;

[0091] Step S306: Record the learning behavior, rest behavior, and interaction behavior in the first set of data as 00, 01, and 10 respectively, and sort them according to the corresponding date and collection time to form the time sequence data of personnel behavior; the encoding facilitates rapid comparison and segmentation in subsequent algorithms.

[0092] Step S307: For any two adjacent data in the time series data of personnel behavior for any date, if the two data are different, the middle of the two data is taken as the dividing position. All dividing positions are obtained repeatedly and the data is divided to obtain the first behavior sequence.

[0093] Step S308: For any two adjacent division positions in the first behavior sequence, mark them as corresponding behavior time periods according to the data contained in the two division positions. Behavior time periods include learning time periods, rest time periods, and interaction time periods. For example, if the data contained in two division positions is 00, then the time period corresponding to the two division positions is the learning time period.

[0094] Step S309: For any behavior time period, denoted as the first behavior time period, obtain the illuminance level of the first energy-saving lamp corresponding to each collection time in the first behavior time period, denoted as the first light intensity sequence;

[0095] Step S310: Obtain the proportion of each light intensity level in the first light intensity sequence, and obtain the light intensity level with the largest proportion, which is recorded as the first light intensity level. At the same time, obtain the indoor light intensity level collected at the sampling time corresponding to each light intensity level with the largest proportion, and calculate the average value, which is recorded as the indoor reference light intensity. Record the first light intensity level as the best light intensity under the indoor reference light intensity for the first row of the time period. Determine the most energy-efficient illuminance that can meet visual needs directly based on historical real usage data.

[0096] Step S311: Repeatedly acquire the indoor reference light intensity and optimal light intensity corresponding to all time periods in the first row sequence, and record them as the optimal light intensity data for the corresponding date of the first row sequence; and repeatedly acquire the optimal light intensity data for all dates.

[0097] In the specific implementation process, for any two adjacent division positions in the first row sequence, if the two adjacent division positions contain too little data, for example, only one or two, the interval can be removed and not analyzed further, because if the data contained is too little, it may be abnormal data caused by sensor fluctuations and has no reference value.

[0098] Step S4 involves controlling and adjusting the energy-saving lamps based on the predicted personnel presence data and the optimal light intensity data. Step S4 includes the following sub-steps:

[0099] For step S401, please refer to... Figure 3 As shown, for the current moment, the time period corresponding to the personnel presence prediction data is recorded as the current time period; if there are no people in the first lighting range in the current time period, and the next time period adjacent to the current time period in the personnel presence prediction data is an unoccupied time period, then the first energy-saving lamp is turned off; to avoid turning off the lamp too early due to short-term prediction errors, which would affect the user experience, and to only cut off the power when there is no expected continuous unoccupied status, so as to minimize the no-load consumption.

[0100] Step S402: If there are people in the first lighting range during the current time period, or if the next time period adjacent to the current time period in the prediction data is a time period with people, then turn on the first energy-saving lamp; when it is predicted that there will be people in the next time period, the energy-saving lamp can be turned on in advance to improve the experience of people.

[0101] Step S403: At the current moment, with the first energy-saving lamp on, obtain the indoor light intensity and record it as the current indoor light intensity HD; if there are people within the first lighting range, obtain the corresponding behavior type at the current moment and record it as the current behavior; if there are no people within the first lighting range, record the learned behavior as the current behavior; the indoor light intensity serves as the basis for adjusting the energy-saving lamp light intensity. For example, if the indoor light intensity is greater than the optimal light intensity during the day, the energy-saving lamp does not need to provide additional light intensity.

[0102] Step S404: Obtain the best light intensity data within k4 dates prior to the current time, and record it as the reference best light intensity data; from the reference best light intensity data, obtain the two best light intensities with the smallest difference from the current indoor light intensity for the current behavior time period, and record them as LA and LB respectively, and record the indoor reference light intensities corresponding to LA and LB as HA and HB respectively in order, where k4 is the number of settings; in this embodiment, k4=7; different behaviors and different indoor light intensities have significantly different light requirements, and the corresponding best light intensity is selected according to the behavior type and indoor light intensity;

[0103] Step S405: Calculate LZ and adjust the illuminance of the first energy-saving lamp to LZ, where LZ = Q3 * LA + Q4 * LB. If HB - HD = HA - HD = 0, then Q3 = Q4 = 0.5; otherwise, Q3 = |HB - HD| / (|HB - HD| + |HA - HD|), Q4 = 1 - Q3. An average is calculated based on the closeness of the current conditions to the reference conditions to avoid sudden large jumps and improve visual comfort.

[0104] Step S406: Repeat the control of all energy-saving lamps in the classroom;

[0105] In practice, each energy-saving lamp can be controlled individually, allowing for dynamic adjustment of the lighting area and brightness based on personnel location and activity area. This reduces energy consumption and extends the lifespan of lighting equipment.

[0106] Example 2, please refer to Figure 4 As shown, Figure 4 A schematic diagram of an electronic device is provided, which may include a processor, a communication interface, a memory, and a communication bus. The processor, communication interface, and memory communicate with each other via the communication bus. The memory stores computer-readable instructions, and the processor can call these instructions. When the processor executes a computer-readable instruction, it performs steps such as those in an AI-based energy-saving lamp control method to achieve the following functions: deploying various sensors in the classroom to collect data on the presence and behavior of people within the illumination range of the energy-saving lamp, and collecting light intensity control data for the energy-saving lamp; constructing a predictive model based on the presence data and performing predictive analysis to obtain predicted presence data; analyzing light intensity based on the behavior data and light intensity control data to obtain optimal light intensity data; and controlling and adjusting the energy-saving lamp according to the predicted presence data and the optimal light intensity data.

[0107] Furthermore, the logical instructions in the aforementioned memory can be implemented as software functional units and sold or used as independent products, and can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0108] Example 3: This application also provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it performs steps such as those in the AI-based energy-saving lamp control method to achieve the following functions: deploying various sensors in the classroom to collect data on the presence and behavior of people within the illumination range of the energy-saving lamp, and collecting light intensity control data of the energy-saving lamp; constructing a prediction model based on the presence data and performing presence prediction analysis to obtain presence prediction data; performing light intensity analysis based on the behavior data and light intensity control data to obtain optimal light intensity data; and controlling and adjusting the energy-saving lamp according to the presence prediction data and the optimal light intensity data.

[0109] Based on the above description of the embodiments, the embodiments of the present invention can be provided as methods, systems, or computer program products. Based on this understanding, the above technical solutions, in essence or in terms of their contribution to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or certain parts of the embodiments.

[0110] In the embodiments provided in this application, it should be understood that the disclosed system or method can be implemented in other ways. The embodiments described above are merely illustrative. For example, the division of modules or units is only a logical functional division, and there may be other division methods in actual implementation. Furthermore, multiple modules or units may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the coupling or direct coupling or communication connection shown or discussed may be through some communication interfaces. The indirect coupling or communication connection between systems, modules, and units may be electrical, mechanical, or other forms.

[0111] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. An AI-based energy-saving lamp control method, characterized in that, Includes the following steps: Multiple sensors are installed in the classroom to collect data on the presence and behavior of people within the illumination range of the energy-saving lamps, as well as data on the light intensity control of the energy-saving lamps. A predictive model is built based on personnel presence data, and personnel presence prediction analysis is performed to obtain personnel presence prediction data; Based on human behavior data and light intensity control data, light intensity analysis was performed to obtain the optimal light intensity data; The energy-saving lamps are controlled and adjusted based on the predicted personnel presence data and the optimal light intensity data; Based on human behavior data and light intensity control data, light intensity analysis is performed to obtain optimal light intensity data, including the following sub-steps: For any collection time in the personnel behavior data, it is recorded as the first time. The total number of personnel collected at the first time is obtained, and the proportion of personnel with learning behavior, rest behavior and interactive behavior at the first time is calculated respectively and recorded as PA1, PA2 and PA3 in order. If PA1=PA2=PA3, then mark the first moment as the learning behavior; Otherwise, select the two largest values ​​among PA1, PA2, and PA3, and label them VA1 and VA2 respectively in descending order; If VA1 equals VA2, and VA1 and VA2 include PA1, then the first moment is marked as a learning behavior; otherwise, the first moment is marked as an interactive behavior. The process of analyzing light intensity based on human behavior data and light intensity control data to obtain optimal light intensity data also includes the following sub-steps: If VA1 is not equal to VA2, and VA1 and VA2 include PA1 and PA1 > k3*(VA2+VA1), then the first time step is marked as the learning behavior; otherwise, the first time step is marked as the behavior type corresponding to max(VA1, VA2), where k3 is the set scaling factor, k3≤0.5; Repeatedly mark all collection times in the personnel behavior data to obtain the first behavior data; The learning behavior, rest behavior, and interaction behavior in the first set of data are recorded as 00, 01, and 10 respectively, and sorted according to the corresponding date and collection time to form the time series data of personnel behavior.

2. The AI-based energy-saving lamp control method according to claim 1, characterized in that, The process of deploying various sensors in the classroom to collect data on the presence and behavior of people within the illumination range of energy-saving lamps, and to collect light intensity control data for the energy-saving lamps, includes the following sub-steps: Human presence sensors, light intensity sensors, and image sensors are deployed in the classroom; any energy-saving lamp in the classroom is designated as the first energy-saving lamp, and the illumination range of the first energy-saving lamp is designated as the first illumination range. Obtain the theoretical range of indoor light intensity and divide it into k1 indoor light intensity levels, where k1 is the number of levels set; A human presence sensor is used to collect data on the presence of people within a first lighting range at a first time interval, and the time of collection is recorded. This data is denoted as the human presence data of the first energy-saving lamp, where the first time interval is t1.

3. The AI-based energy-saving lamp control method according to claim 2, characterized in that, The process of deploying various sensors in the classroom to collect data on the presence and behavior of people within the illumination range of energy-saving lamps, as well as data on the light intensity control of the energy-saving lamps, also includes the following sub-steps: If there are people within the first lighting range, the image sensor is used to collect behavioral images of each person within the first lighting range at a first time interval, and synchronously with the human presence sensor. The behavioral image of each person is analyzed to obtain the behavioral type of each person, which is recorded as the human behavior data of the first energy-saving lamp. The behavioral types of the people include learning behavior, resting behavior and interactive behavior. If there are people within the first lighting range, the illuminance level of the first energy-saving lamp and the indoor light intensity level are collected synchronously with the human presence sensor at a first time interval and recorded as light intensity control data.

4. The AI-based energy-saving lamp control method according to claim 3, characterized in that, The process of building a predictive model based on personnel presence data and performing personnel presence prediction analysis to obtain personnel presence prediction data includes the following sub-steps: The presence of personnel within the first lighting range in the personnel presence data is recorded as 1, and the absence of personnel is recorded as 0. The data is then sorted according to the corresponding date and collection time, and divided according to the corresponding date to obtain the personnel presence time series data. For any given date, the time sequence data of personnel is arranged in ascending order of time and denoted as the first personnel existence sequence; the second time period is set to ZT2. The sequence of the first person is divided into multiple subsequences with a time length of ZT2 using ZT2. Any subsequence is denoted as the first subsequence, and the time period corresponding to the first subsequence is denoted as the first time period.

5. The AI-based energy-saving lamp control method according to claim 4, characterized in that, Building a predictive model based on personnel presence data and performing personnel presence predictive analysis to obtain personnel presence prediction data also includes the following sub-steps: For the first subsequence, obtain the proportion coefficient of 1 in the first subsequence, and denot it as the personnel coefficient of the first subsequence; repeatedly obtain the personnel coefficients of all subsequences in the first personnel existence sequence, and arrange them according to the corresponding time order, and denot them as the personnel coefficient sequence of the first personnel existence sequence; Repeatedly obtain the personnel coefficient sequence for all dates in the personnel data, and concatenate them sequentially according to date order, denoted as the first personnel coefficient data; For the first subsequence, obtain the subsequence corresponding to the first time period in other dates, arrange them in date order, and record them as the personnel coefficient data of the first time period. Repeat this process to obtain the personnel coefficient data of all time periods and record them as the second personnel coefficient data.

6. The AI-based energy-saving lamp control method according to claim 5, characterized in that, Building a predictive model based on personnel presence data and performing personnel presence prediction analysis to obtain personnel presence prediction data also includes the following sub-steps: The Long Short-Term Memory Network was trained using the first and second personnel coefficient data, respectively, and the first and second prediction models were obtained after completion. The mean absolute errors of the first prediction model and the second prediction model are obtained respectively and denoted as MA1 and MA2 in order. Using the first personnel coefficient data and the first prediction model, the personnel coefficient data for all time periods over the next k2 dates are predicted to obtain the first prediction data, where k2 is the set number of dates; Predict the personnel coefficient data for each time period in the next k2 dates using the second personnel coefficient data and the second prediction model, and arrange the personnel coefficient data for each time period predicted for each future date in the corresponding chronological order to obtain the personnel coefficient data for each future date, denoted as the second prediction data; For any time period in the next k2 dates, denoted as the second time period; respectively obtain the corresponding data of the first prediction data and the second prediction data in the second time period, and denote them as AE1 and AE2 in sequence; Calculate AE0, denoted as the final personnel coefficient for the second time period, where AE0 = Q1 * AE1 + Q2 * AE2, Q1 = MA2 / (MA1 + MA2), Q2 = MA1 / (MA1 + MA2); and calculate MA0, MA0 = Q1 * MA1 + Q2 * MA2; For the second time period, if AE0 < MA0, mark the second time period as an unoccupied period, otherwise mark the second time period as an occupied period. Repeat the marking for all time periods in the next k2 dates. After completion, obtain the personnel presence prediction data.

7. The AI-based energy-saving lamp control method according to claim 6, characterized in that, Based on the personnel behavior data and the light intensity control data for light intensity analysis, obtaining the optimal light intensity data further includes the following sub-steps: For any two adjacent data in the personnel behavior time series data for any date, if the two data are different, then the middle of the two data is used as the division position. Repeat to obtain all division positions and perform division to obtain the first behavior sequence; For any two adjacent division positions in the first behavior sequence, mark them as the corresponding behavior periods according to the data included in the two division positions. The behavior periods include study periods, rest periods, and interaction periods; For any behavior period, denoted as the first behavior period, obtain the light intensity level of the first energy-saving lamp corresponding to each collection moment in the first behavior period, denoted as the first light intensity sequence; Obtain the proportion of each light intensity level in the first light intensity sequence, and obtain the light intensity level with the largest proportion, denoted as the first light intensity level. At the same time, obtain the indoor light intensity level collected at the collection moment corresponding to each light intensity level with the largest proportion, and calculate the average value, denoted as the indoor reference light intensity; Denote the first light intensity level as the optimal light intensity under the indoor reference light intensity for the first behavior period; Repeat to obtain the indoor reference light intensity and the optimal light intensity corresponding to all behavior periods in the first behavior sequence, denoted as the optimal light intensity data for the date corresponding to the first behavior sequence; and repeat to obtain the optimal light intensity data for all dates.

8. The AI-based energy-saving lamp control method according to claim 7, characterized in that, Controlling and adjusting the energy-saving lamp according to the personnel presence prediction data and the optimal light intensity data includes the following sub-steps: For the current moment, denote the time period corresponding to the current moment in the personnel presence prediction data as the current time period; if there is no person in the first lighting range of the current time period and the next time period adjacent to the current time period in the personnel presence prediction data is an unoccupied period, then turn off the first energy-saving lamp; If there are people within the first lighting range during the current time period, or if the next time period adjacent to the current time period in the prediction data is a time period with people, then turn on the first energy-saving lamp; At the current moment, with the first energy-saving lamp on, the indoor light intensity is obtained and recorded as the current indoor light intensity HD; If there are people within the first lighting range, obtain the corresponding behavior type at the current moment and record it as the current behavior; if there are no people within the first lighting range, record the learned behavior as the current behavior. Obtain the best light intensity data within k4 dates prior to the current time, and record it as the reference best light intensity data; The two best light intensities with the smallest difference from the current indoor light intensity are obtained from the reference best light intensity data for the current behavior period, and are denoted as LA and LB respectively. The indoor reference light intensities corresponding to LA and LB are denoted as HA and HB respectively in order, where k4 is the number of settings. Calculate LZ and adjust the illuminance of the first energy-saving lamp to LZ, where LZ = Q3 * LA + Q4 * LB, where if HB - HD = HA - HD = 0, then Q3 = Q4 = 0.5, otherwise Q3 = |HB - HD| / (|HB - HD| + |HA - HD|), Q4 = 1 - Q3; Repeatedly control all energy-saving lights in the classroom.

Images