Intelligent lamp control system based on optics and radar perception

By combining multiple sensors and machine learning algorithms, the intelligent lamp control system can accurately identify activity types and light changes in complex environments, solving the problem of misoperation and improving user experience and system adaptability.

CN118984515AInactive Publication Date: 2024-11-19SHIJIAZHUANG ZHONGXIN RUIERWEI TECH DEV

Patent Information

Application Number
CN202410843684.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-06-27
Publication Date
2024-11-19
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The existing intelligent lamp control system based on optical and radar perception is prone to false alarms and misoperation in complex dynamic environments, resulting in poor user experience, especially when pets, children's activities and external light changes, the system is difficult to accurately distinguish the activity types. Causes frequent adjustment of brightness or switching state.

Method used

Using perception modules combined with optical, radar, infrared and ultrasonic sensors, data fusion and classification are carried out through particle filtering algorithms and SVM-CNN models, to identify the activity types of humans, pets and children, and smoothly process them according to changes in ambient light to generate optimized Lamp control strategy, dynamically adjust lighting settings.

Benefits of technology

It realizes accurate identification and distinction between different activity types, reduces misoperation, improves the system's resource utilization efficiency and user experience, and provides a more comfortable and intelligent lighting effect.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118984515A_ABST
    Figure CN118984515A_ABST
Patent Text Reader

Abstract

The invention discloses an intelligent lamp control system based on optics and radar perception, and belongs to the technical field of lamp control systems. The method and the device are used for solving the problem of poor user experience caused by frequent false alarm or misoperation of a system in an environment with a large number of moving objects or rapid light change in an existing scheme. By collecting and marking activity data of human, pets and children, the system can accurately identify various activity types, such as walking and standing of the human, running and standing of the pets and playing and jumping of the children. A motion trail, speed and acceleration formula and ambient light change L (t) are used as auxiliary features, the system performs data cleaning and standardization processing, and a training data set and a verification data set are constructed. An SVM-CNN algorithm is selected to generate a model, and different activity types are efficiently classified through spatial feature extraction of a convolutional layer, sampling of a pooling layer and flattening and connection of a full connection layer. After the model is trained and verified, sensor data are analyzed in real time, lamp control is driven, and intelligent dimming and switch control are achieved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of lamp control systems, and in particular to an intelligent lamp control system based on optical and radar perception. Background Art

[0002] After searching, the application number 202311288284.4 discloses an intelligent lamp control system based on optical and radar perception, including a lamp operation analysis unit, a lamp environment status monitoring unit, a processor and a lamp control prediction processing unit. The present invention analyzes the operation status of the intelligent lamp in the past period of time and the operation status in the past period of time, divides the time period into time frequency according to the lamp control in different time periods, obtains and analyzes the relevant operation status of lamps with different time period frequencies, and converts data and signals based on the data of the relevant operation status, and processes the converted data and signals with the environment of the lamp in the current period to obtain relevant control coefficients, controls the intelligent lamps according to the control coefficients, and controls the lamps accurately according to the demand situation, thereby saving power consumption and improving the control efficiency of the intelligent lamps.

[0003] However, in the lighting control system in the comparative document, when pets are walking or jumping around at home, radar sensors and optical sensors may be frequently triggered, causing the system to misjudge human activity and frequently turn the lights on or off. Similarly, when children are running and playing indoors, the sensors will frequently detect movement due to their frequent and large movements. It is difficult for the system to distinguish these activities from normal adult activities, which may cause the lights to constantly change state. In addition, rapid light changes caused by cars passing by or clouds moving outside the window may trigger optical sensors, causing the lights to frequently adjust brightness or switch on and off, thereby interfering with the user's normal life and resulting in a poor user experience. These situations show that existing lighting control systems have obvious problems with false alarms and misoperations in complex dynamic environments. Summary of the invention

[0004] The purpose of the present invention is to provide an intelligent lighting control system based on optical and radar perception, which is used to solve the technical problems of poor results in the analysis and expansion of individual detection data and overall assembly data of battery assembly in existing solutions.

[0005] The purpose of the present invention can be achieved through the following technical solutions:

[0006] An intelligent lighting control system based on optical and radar perception, comprising:

[0007] The perception module is used to detect the ambient light intensity and detect the movement in the environment by transmitting and receiving microwave signals, and finally collect and process the acquired optical data and radar data;

[0008] The processor is responsible for receiving data from the perception module and fusing and processing the obtained data;

[0009] The lamp operation analysis unit is used to collect the operation data of the intelligent lamps, analyze the operation time points, determine the frequent operation time periods and the occasional operation time periods, and optimize the control strategy of the lamps based on the operation analysis results;

[0010] A lamp control prediction processing unit predicts the operating state of the lamp based on the control signal group generated by the processor, executes the switch and brightness adjustment instructions of the lamp, and generates a control signal group according to the environmental state, user behavior and historical data when executing the command;

[0011] An execution module, used to adjust the brightness, color temperature or switch state of the lamp according to the instruction of the lamp control prediction processing unit;

[0012] The power management unit is used to provide a stable power supply to ensure the reliable operation of the system.

[0013] Optionally, the perception module includes an optical sensor, a radar sensor, an infrared sensor, an ultrasonic sensor, and an environmental sensor. i The collected data i,k Assign precise timestamp t i,k , according to the timestamp t i,k For different sensors S i Data i,k Align and interpolate the data with asynchronous sampling time to make the fusion analysis results at the same time point. i (t), fusion analysis results d i The expression of (t) is:

[0014] Establish the spatial mapping relationship between different sensors and transform a sensor S i The data coordinates are converted to another sensor S j In the coordinate system, the data point d in the coordinate transformation is obtained. j ′(t), which is used to obtain the data point d in the coordinate transformation j The expression of ′(t) is: Among them, R ij is the rotation matrix, which transforms the coordinates from the sensor S i The coordinate system of the sensor S is rotated to another j coordinate system, tij is the translation vector, which transforms the coordinates from the sensor S i The coordinate system of the sensor S is translated to another sensor S j The coordinate system of is a 4x1 column vector.

[0015] The spatially synchronized data is fused using a particle filter algorithm to fuse the data from different sensors into a comprehensive perception result D(t). The expression for obtaining the perception result D(t) is:

[0016] in The weight of the i-th particle in the particle filter.

[0017] Optionally, the processor performs the following steps when processing the perception module data:

[0018] Collect data on human, pet and child activities, and mark them separately. Then, based on the motion trajectory T(t) and speed data v(t), further mark human walking, standing, pet running, stillness, and children playing and jumping. The expression for obtaining the motion trajectory T(t) is: T(t) = (x(t), y(t), z(t)), where x(t), y(t), z(t) are the position coordinates of the object at time t.

[0019] The motion trajectory T(t) of the moving object and the change of ambient light L(t) are extracted from the perception result D(t). The typical motion patterns of different activity types are identified by analyzing the trajectory data and the velocity v(t) and acceleration a(t) of the moving object are calculated using the motion trajectory data. The change of ambient light L(t) is used as an auxiliary feature. The expressions of velocity v(t) and acceleration a(t) are:

[0020]

[0021] in, is the first-order derivative of the position vector x(t) with respect to time t.

[0022] The collected data is cleaned and standardized, and the labeled data is classified by activity type to construct a training data set. Each data set contains multiple features, such as motion trajectory, speed, acceleration, and light changes. According to different activity types, the data is divided into multiple subsets, which are used as training sets and validation sets in turn.

[0023] Select the SVM-CNN algorithm to generate the model according to the data characteristics and classification requirements, use the training set to train the SVM-CNN model, and finally use the validation set to validate the trained SVM-CNN model;

[0024] The trained model is deployed to the system, and the sensor data is received in real time for analysis and classification. The model determines the current activity type based on the input features and outputs the classification results. The classification results are used to drive the lighting control system to achieve intelligent dimming and switch control.

[0025] Through online learning and incremental learning, model parameters are continuously adjusted and optimized based on user feedback and actual operation data.

[0026] Optionally, the SVM-CNN model includes a convolutional layer, a pooling layer, and a fully connected layer, wherein the convolutional layer is used to extract spatial features, the pooling layer is used for downsampling, and the fully connected layer is used to flatten and connect the output of the pooling layer. The SVM uses the high-level features extracted by the CNN for classification and adopts a one-to-many strategy in multi-classification problems.

[0027] Optionally, when processing the perception module data, the processor also needs to perform person recognition on the image data captured by the optical sensor, identify each person in the room, assign a unique identifier, and set a corresponding weight according to the characteristics of the person;

[0028] Set adjustment priorities based on weight and activity, and make dynamic adjustments based on the character's activity status and weight.

[0029] Optionally, after receiving the optical information, the perception module will also collect ambient light intensity data in real time through an optical sensor, record the time series of light changes, and when a sensor fails, a backup sensor can immediately take over its work;

[0030] In each time window, the light intensity data is averaged or weighted averaged to reduce the fluctuation caused by instantaneous light changes and perform exponential weighted averaging on the light intensity data. The weight of new data points gradually decreases, and recent data has a greater impact on the results, thereby smoothing the light changes and obtaining smoothed light data.

[0031] According to the smoothed light data, the brightness and switch status of the lamp are adjusted in real time, so that the brightness of the lamp can be adjusted smoothly.

[0032] Use the light change buffer algorithm to smooth light changes and instantaneous light changes.

[0033] Optionally, the steps for collecting the operation data of the intelligent lamps and analyzing the operation time points to determine the frequent operation time periods and the occasional operation time periods, and optimizing the control strategy of the lamps based on the operation analysis results are as follows:

[0034] Integrate data acquisition modules in smart lamps to record the operating status of lamps in real time, including switch status, brightness adjustment, usage time, etc.

[0035] The collected operation data is sorted in chronological order to construct a time series data set. The switch time, usage duration and brightness adjustment of the lamps are analyzed, and the frequency analysis of the time series data is performed to determine the usage frequency of the lamps in different time periods of the day. Identify high-frequency usage periods and low-frequency usage periods;

[0036] Perform frequency analysis on time series data to determine the frequency of lamp usage at different time periods during the day and identify high-frequency and low-frequency usage periods;

[0037] Use data mining technology to analyze users' lighting usage habits in different time periods and identify high-frequency and low-frequency usage patterns;

[0038] Correlate and analyze the lamp operation data with environmental parameters and user behavior data to identify the correlation between specific behaviors and lamp operation;

[0039] Generate optimized lighting control strategies based on the results of operation time point analysis and operation mode recognition.

[0040] Optionally, the lamp control prediction processing unit performs the following steps when performing the prediction control strategy: parsing the received control signal to identify the specific operation to be performed; parsing the received control signal to identify the specific operation to be performed;

[0041] A lamp operation status model is established, taking into account the current environmental status, user behavior and historical data. The model predicts the switch status, brightness level and usage frequency of the lamp, and inputs the current environmental data, user behavior data and historical data characteristics into the model for status prediction. Finally, the model is used to calculate and predict the operating status of the lamp in the future.

[0042] Optionally, the lamp operation analysis unit further includes the following steps when adjusting the control strategy:

[0043] Monitor the current environmental status and sensor working status, environmental changes and user behavior in real time, and dynamically adjust the sensitivity of the sensor based on real-time environmental data;

[0044] When the system detects a degradation in sensor performance or a drastic change in the environment, it immediately issues a warning signal and adjusts the control strategy.

[0045] Compared with the existing solutions, the present invention achieves the following beneficial effects:

[0046] By collecting and labeling the activity data of humans, pets and children, the system can accurately identify various types of activities, such as human walking and standing, pets running and standing still, and children playing and jumping. Using the motion trajectory, speed and acceleration formula, and ambient light change L(t) as auxiliary features, the system performs data cleaning and standardization processing to construct training data sets and verification data sets. The SVM-CNN algorithm is used to generate the model, and the spatial features are extracted through the convolution layer, the pooling layer is downsampled, and the fully connected layer is flattened and connected to efficiently classify different activity types. After training and verification, the model analyzes sensor data in real time, drives lamp control, and realizes intelligent dimming and switch control.

[0047] When the processor processes the data of the perception module, the present invention performs character recognition on the image data captured by the optical sensor. The system can identify each person in the room and assign a unique identifier, and set corresponding weights based on the characteristics of the person. Accurate recognition and weight allocation enable the system to adjust the lighting control in a personalized manner. The adjustment priority is set according to the weight and activity status, and dynamic adjustments are made according to the activity status and weight of the person, so that the system can respond to user needs and environmental changes in real time. This not only improves the resource utilization efficiency of the system and reduces misoperation, but also provides a more comfortable and intelligent lighting experience, improving user satisfaction and system adaptability.

[0048] The present invention integrates a data acquisition module in the intelligent lamp to record the operating status of the lamp (including switch status, brightness adjustment, usage time, etc.) in real time, and the system can build a detailed time series data set. By analyzing these data, the system can identify the frequency of lamp use in different time periods of the day and identify high-frequency and low-frequency usage periods. Using data mining technology, the system can analyze the user's usage habits in different time periods and identify high-frequency and low-frequency usage patterns. By correlating and analyzing the lamp operation data, environmental parameters and user behavior data, the system can identify the relationship between specific behaviors and lamp operation. Based on these analyses, the system generates an optimized lamp control strategy, and in the lamp control prediction processing unit, parses the received control signal, integrates the environmental status, user behavior and historical data, and predicts the future lamp operation status. These measures improve the intelligence level, energy efficiency and user experience of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The present invention will be further described below in conjunction with the accompanying drawings.

[0050] Figure 1 This is a module block diagram of an intelligent lighting control system based on optical and radar perception according to the present invention. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary operation and maintenance personnel in this field without creative work are within the scope of protection of the present invention.

[0052] like Figure 1 As shown, the present invention is an intelligent lighting control system based on optical and radar perception, including a perception module for detecting ambient light intensity, and detecting motion in the environment by transmitting and receiving microwave signals, and finally collecting and processing the acquired optical data and radar data;

[0053] The processor is responsible for receiving data from the perception module and fusing and processing the obtained data;

[0054] The lamp operation analysis unit is used to collect the operation data of the intelligent lamps, analyze the operation time points, determine the frequent operation time periods and the occasional operation time periods, and optimize the control strategy of the lamps based on the operation analysis results;

[0055] A lamp control prediction processing unit predicts the operating state of the lamp based on the control signal group generated by the processor, executes the switch and brightness adjustment instructions of the lamp, and generates a control signal group according to the environmental state, user behavior and historical data when executing the command;

[0056] An execution module, used to adjust the brightness, color temperature or switch state of the lamp according to the instruction of the lamp control prediction processing unit;

[0057] The power management unit is used to provide a stable power supply to ensure the reliable operation of the system.

[0058] In this embodiment, the perception module includes an optical sensor, a radar sensor, an infrared sensor, an ultrasonic sensor, and an environmental sensor. i The collected data i,k Assign precise timestamp t i,k , according to the timestamp t i,k For different sensors S i Data i,k Align and interpolate the data with asynchronous sampling time to make the fusion analysis results at the same time point. i (t), fusion analysis results d i The expression of (t) is:

[0059] Establish the spatial mapping relationship between different sensors and transform a sensor S iThe data coordinates are converted to another sensor S j In the coordinate system, the data point d in the coordinate transformation is obtained. j ′(t), which is used to obtain the data point d in the coordinate transformation j The expression of ′(t) is: Among them, R ij is the rotation matrix, which transforms the coordinates from the sensor S i The coordinate system of the sensor S is rotated to another j coordinate system, tij is the translation vector, which transforms the coordinates from the sensor S i The coordinate system of the sensor S is translated to another sensor S j The coordinate system of is a 4x1 column vector.

[0060] The spatially synchronized data is fused using a particle filter algorithm to fuse the data from different sensors into a comprehensive perception result D(t). The expression for obtaining the perception result D(t) is:

[0061] in The weight of the i-th particle in the particle filter.

[0062] It should be noted that by assigning accurate timestamps to the data collected by each sensor and performing time alignment and interpolation processing on the data of different sensors, all data can be fused and analyzed at the same time point. This method can solve the problem of asynchronous sensor sampling time and improve the accuracy and consistency of data fusion.

[0063] By establishing a spatial mapping relationship between different sensors and using a coordinate transformation matrix to transform the data coordinates of one sensor into the coordinate system of another sensor, it is ensured that the data of all sensors are processed in the same spatial reference frame. This can eliminate the spatial errors caused by different sensor installation positions and directions and improve the accuracy of spatial data fusion.

[0064] The particle filter algorithm is used to fuse the data of different sensors into a comprehensive perception result, which can effectively process multi-source heterogeneous data and improve the comprehensive perception ability of the system. The particle filter algorithm integrates the data of different sensors through the weighted average method, which can better cope with environmental changes and noise interference, and improve the robustness and reliability of the perception results.

[0065] The combination of precise time and space synchronization, multi-sensor data fusion and particle filter algorithm enables the system to process and respond to environmental changes in real time and improve the system's dynamic adaptability. For example, under complex environmental conditions, the system can quickly adjust the perception and control strategies to provide accurate and reliable perception results to meet the needs of different scenarios.

[0066] For example: the optical sensor is installed in the center of the room, and the radar sensor is installed in the corner of the room. The data coordinates of the radar sensor are converted to the coordinate system of the optical sensor through the rotation matrix Rij and the translation vector tij to ensure that all data are processed in the same spatial reference frame. The optical sensor records image data once per second, the radar sensor records motion data 10 times per second, and the infrared sensor records heat source data 5 times per second. Through a unified time base, the data timestamps of all sensors are ensured to be consistent. The image data provided by the optical sensor is used to identify objects and positions in the room, the motion data provided by the radar sensor is used to track the movement trajectory of objects, and the heat source data provided by the infrared sensor is used to detect human heat sources. The particle filter algorithm weights and averages these data to generate a comprehensive perception result. When someone enters the room, the system automatically turns on the lights and adjusts the brightness; when no one is detected in the room, the system automatically turns off the lights to save energy. Through precise perception and real-time response, the smart home system can provide a more comfortable, energy-saving and safe living environment.

[0067] In this embodiment, the processor performs the following steps when processing the perception module data:

[0068] Collect data on human, pet and child activities, and mark them separately. Then, based on the motion trajectory T(t) and speed data v(t), further mark human walking, standing, pet running, stillness, and children playing and jumping. The expression for obtaining the motion trajectory T(t) is: T(t) = (x(t), y(t), z(t)), where x(t), y(t), z(t) are the position coordinates of the object at time t.

[0069] The motion trajectory T(t) of the moving object and the change of ambient light L(t) are extracted from the perception result D(t). The typical motion patterns of different activity types are identified by analyzing the trajectory data and the velocity v(t) and acceleration a(t) of the moving object are calculated using the motion trajectory data. The change of ambient light L(t) is used as an auxiliary feature. The expressions of velocity v(t) and acceleration a(t) are:

[0070]

[0071] in, is the first-order derivative of the position vector x(t) with respect to time t.

[0072] The collected data is cleaned and standardized, and the labeled data is classified by activity type to construct a training data set. Each data set contains multiple features, such as motion trajectory, speed, acceleration, and light changes. According to different activity types, the data is divided into multiple subsets, which are used as training sets and validation sets in turn.

[0073] Select the SVM-CNN algorithm to generate the model according to the data characteristics and classification requirements, use the training set to train the SVM-CNN model, and finally use the validation set to validate the trained SVM-CNN model;

[0074] The trained model is deployed to the system, and the sensor data is received in real time for analysis and classification. The model determines the current activity type based on the input features and outputs the classification results. The classification results are used to drive the lighting control system to achieve intelligent dimming and switch control.

[0075] Through online learning and incremental learning, model parameters are continuously adjusted and optimized based on user feedback and actual operation data.

[0076] In this embodiment, the SVM-CNN model includes a convolutional layer, a pooling layer and a fully connected layer, wherein the convolutional layer is used to extract spatial features, the pooling layer is used for downsampling, and the fully connected layer is used to flatten and connect the output of the pooling layer. The SVM uses the high-level features extracted by CNN for classification and adopts a one-to-many strategy in multi-classification problems.

[0077] It should be noted that the convolution layer can detect the outline of people in the room, the shape of pets, and the details of children's movements, which are important features for activity classification. The pooling layer can simplify the detected edge and shape features, making the subsequent fully connected layer processing more efficient. The fully connected layer can combine the different activity features in the room to form a comprehensive feature vector to represent the current activity status. The system trains multiple binary classifiers such as "human activity vs. non-human activity", "pet activity vs. non-pet activity", and "children's activity vs. non-children's activity". When human activity is detected, the system can trigger an alarm or record a video; when pet activity is detected, the system can ignore or adjust the sensitivity; when child activity is detected, the system can send a notification to the parent.

[0078] In this embodiment, when processing the perception module data, the processor also needs to perform person recognition on the image data captured by the optical sensor, identify each person in the room, assign a unique identifier, and set a corresponding weight according to the characteristics of the person;.

[0079] Set adjustment priorities based on weight and activity, and make dynamic adjustments based on the character's activity status and weight.

[0080] By performing character recognition on the image data captured by the optical sensor, the system can accurately identify each person in the room and assign a unique identifier to each person. Accurate recognition enables the system to set corresponding weights based on the height, weight, age, etc. of each person, thereby achieving personalized lighting control. By dynamically adjusting the lighting settings, the system can provide the best lighting effect in different situations and improve the user experience.

[0081] For example: In a smart home system, family members include parents and two children. The system monitors the activity status and location of each family member in real time through the perception module and optical sensor, and dynamically adjusts the light settings according to the preset weight and activity status and assigns the father's identifier to ID1, the mother's identifier to ID2, the child 1's identifier to ID3, and the child 2's identifier to ID4. According to the characteristics of each family member (such as age, activity needs), the system sets a weight for each person. Parents have a higher weight because their activities (such as work and reading) require more light; children have a lower weight because their activities (such as playing and resting) require more flexible light. The perception module monitors the activity status and location of each family member in real time. For example, the system detects that the father (ID1) is working in the study, the mother (ID2) is cooking in the kitchen, the child 1 (ID3) is playing in the living room, and the child 2 (ID4) is resting in the bedroom. The system recognizes that the father is working (high weight), the mother is cooking (medium weight), the child 1 is playing (medium weight), and the child 2 is resting (low weight).

[0082] The system sets the adjustment priority according to the activity status and weight of each family member. The father's work activity has the highest priority, the mother's cooking activity has the second highest priority, and the child 1's play activity and the child 2's rest activity have the lowest priority.

[0083] The system dynamically adjusts the lighting settings based on the adjustment priority and real-time activity status. For example, when the father moves from the study to the living room, the system automatically adjusts the lighting settings in the study and living room to ensure that the father can also enjoy appropriate lighting in the living room.

[0084] If the system detects that child 1 enters the bedroom from the living room to prepare for rest, the system will automatically dim the lights in the bedroom according to the weight and activity status to create an environment suitable for rest.

[0085] In this embodiment, after the perception module receives the optical information, it will also collect the ambient light intensity data in real time through the optical sensor, record the time series of light changes, and when a sensor fails, the backup sensor can immediately take over its work;

[0086] In each time window, the light intensity data is averaged or weighted averaged to reduce the fluctuation caused by instantaneous light changes and perform exponential weighted averaging on the light intensity data. The weight of new data points gradually decreases, and recent data has a greater impact on the results, thereby smoothing the light changes and obtaining smoothed light data.

[0087] According to the smoothed light data, the brightness and switch status of the lamp are adjusted in real time, so that the brightness of the lamp can be adjusted smoothly.

[0088] Use the light change buffer algorithm to smooth light changes and instantaneous light changes.

[0089] In this embodiment, the steps for collecting the operation data of the intelligent lamps, analyzing the operation time points, determining the frequent operation time periods and the occasional operation time periods, and optimizing the control strategy of the lamps based on the operation analysis results are as follows:

[0090] Integrate data acquisition modules in smart lamps to record the operating status of lamps in real time, including switch status, brightness adjustment, usage time, etc.

[0091] The collected operation data is sorted in chronological order to construct a time series data set. The switch time, usage duration and brightness adjustment of the lamps are analyzed, and the frequency analysis of the time series data is performed to determine the usage frequency of the lamps in different time periods of the day. Identify high-frequency usage periods and low-frequency usage periods;

[0092] Perform frequency analysis on time series data to determine the frequency of lamp usage at different time periods during the day and identify high-frequency and low-frequency usage periods;

[0093] Use data mining technology to analyze users' lighting usage habits in different time periods and identify high-frequency and low-frequency usage patterns;

[0094] Correlate and analyze the lamp operation data with environmental parameters and user behavior data to identify the correlation between specific behaviors and lamp operation;

[0095] Generate optimized lighting control strategies based on the results of operation time point analysis and operation mode recognition.

[0096] It should be noted that by averaging or weighted averaging the light intensity data in each time window, the system can effectively reduce the fluctuations caused by instantaneous light changes and avoid misoperation caused by sudden light changes. Exponentially weighted averaging of light intensity data gradually reduces the weight of new data points, and recent data has a greater impact on the results, thereby smoothing light changes. Smoothing can enable the system to better cope with gradual changes in ambient light, avoid drastic fluctuations, and improve the stability of lamp brightness adjustment.

[0097] For example: In a smart home system, the perception module collects ambient light intensity data in real time through optical sensors and records the time series of light changes. The system has designed a smoothing and dynamic adjustment mechanism to ensure the stability of the lamp and the comfort of the user. The system collects ambient light intensity data in the room in real time through optical sensors and records the time series of light changes. If an optical sensor fails, the backup sensor will immediately take over its work to ensure uninterrupted data collection. The system performs sliding window processing on the light intensity data, and averages or weighted averages the light data in each time window. For example, every 5 seconds is a window, and the data points in the window are averaged. Based on the smoothed light data, the system adjusts the brightness and switch status of the lamp in real time. For example, when the light intensity gradually decreases, the system gradually increases the brightness of the lamp to avoid sudden brightness changes. Through smoothing and real-time adjustment, the system avoids frequent lamp switching and brightness changes, reduces the user's visual discomfort, and improves the overall user experience. In actual applications, the system detects that the light intensity gradually decreases in the evening. By smoothing the light data, the system increases the brightness of the lamps by 10% every minute to ensure a smooth transition of indoor lighting and provide a comfortable lighting effect.

[0098] In this embodiment, the lamp control prediction processing unit performs the following steps when performing the prediction control strategy: parsing the received control signal, identifying the specific operation to be performed; parsing the received control signal, identifying the specific operation to be performed;

[0099] Establish a lamp operation status model, comprehensively consider the current environmental status, user behavior and historical data; the model predicts the switch status, brightness level and usage frequency of the lamp, and inputs the current environmental data, user behavior data and historical data characteristics into the model to make status predictions; finally, through model calculations, predict the operating status of the lamp in the future. At night, when the user enters a resting state, the system gradually reduces the brightness of the lamp, saving energy while creating a suitable resting environment.

[0100] In this embodiment, the lamp operation analysis unit further includes the following steps when adjusting the control strategy:

[0101] Monitor the current environmental status and sensor working status, environmental changes and user behavior in real time, and dynamically adjust the sensitivity of the sensor based on real-time environmental data;

[0102] When the system detects a degradation in sensor performance or a drastic change in the environment, it immediately issues a warning signal and adjusts the control strategy.

[0103] For example: In a family, the system detects that the user enters the living room at 9 pm, and the living room lights are automatically adjusted to an appropriate brightness. Then, the user enters the bedroom to rest, and the system gradually dims the bedroom lights based on real-time monitoring data to create a comfortable resting environment. Suddenly, the system detects that the external light has decreased sharply (for example, the weather has turned cloudy), and the optical sensor cannot provide accurate data. The system sends out an early warning signal and enables backup sensors (such as infrared sensors) to ensure the accuracy of indoor light adjustment. Through dynamic adjustment and early warning mechanisms, the system always provides stable and accurate lighting services, avoids misoperation caused by light changes or sensor failures, and improves the quality of life of users.

[0104] In addition, the formulas involved in the above are all dimensionless and numerical calculations. They are a formula that is closest to the actual situation obtained by collecting a large amount of data and simulating it with simulation software.

[0105] In the several embodiments provided by the present invention, it should be understood that the disclosed system can be implemented in other ways. For example, the above-described embodiments of the invention are only illustrative, for example, the division of modules is only a logical function division, and there may be other division methods in actual implementation.

[0106] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, and may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0107] In addition, each functional module in each embodiment of the present invention may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0108] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the basic features of the present invention.

[0109] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, ordinary operation and maintenance personnel in the field should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. An intelligent lighting control system based on optical and radar perception, characterized in that: include: The perception module is used to detect the ambient light intensity and detect the movement in the environment by transmitting and receiving microwave signals, and finally collect and process the acquired optical data and radar data; The processor is responsible for receiving data from the perception module and fusing and processing the obtained data; The lamp operation analysis unit is used to collect the operation data of the intelligent lamps, analyze the operation time points, determine the frequent operation time periods and the occasional operation time periods, and optimize the control strategy of the lamps based on the operation analysis results; A lamp control prediction processing unit predicts the operating state of the lamp based on the control signal group generated by the processor, executes the switch and brightness adjustment instructions of the lamp, and generates a control signal group according to the environmental state, user behavior and historical data when executing the command; An execution module, used to adjust the brightness, color temperature or switch state of the lamp according to the instruction of the lamp control prediction processing unit; The power management unit is used to provide a stable power supply to ensure the reliable operation of the system.

2. According to claim 2, the intelligent lighting control system based on optical and radar perception is characterized in that: The perception module includes optical sensors, radar sensors, infrared sensors, ultrasonic sensors, and environmental sensors. i The collected data i,k Assign precise timestamp t i,k , according to the timestamp t i,k For different sensors S i Data i,k Align and interpolate the data with asynchronous sampling time to make the fusion analysis results at the same time point. i (t), fusion analysis results d i The expression of (t) is: Establish the spatial mapping relationship between different sensors and transform a sensor S i The data coordinates are converted to another sensor S j In the coordinate system, the data point d in the coordinate transformation is obtained. j ′(t), which is used to obtain the data point d in the coordinate transformation j The expression of ′(t) is: Among them, R ij is the rotation matrix, which transforms the coordinates from the sensor S i The coordinate system of the sensor S is rotated to another j coordinate system, tij is the translation vector, which transforms the coordinates from the sensor S i The coordinate system of the sensor S is translated to another sensor S j The coordinate system of is a 4x1 column vector. The spatially synchronized data is fused using a particle filter algorithm to fuse the data from different sensors into a comprehensive perception result D(t). The expression for obtaining the perception result D(t) is: in The weight of the i-th particle in the particle filter.

3. The intelligent lighting control system based on optical and radar perception according to claim 2, characterized in that: The processor performs the following steps when processing the perception module data: Collect data on human, pet and child activities, and mark them separately. Then, based on the motion trajectory T(t) and speed data v(t), further mark human walking, standing, pet running, stillness, and children playing and jumping. The expression for obtaining the motion trajectory T(t) is: T(t) = (x(t), y(t), z(t)), where x(t), y(t), z(t) are the position coordinates of the object at time t. The motion trajectory T(t) of the moving object and the change of ambient light L(t) are extracted from the perception result D(t). The typical motion patterns of different activity types are identified by analyzing the trajectory data and the velocity v(t) and acceleration a(t) of the moving object are calculated using the motion trajectory data. The change of ambient light L(t) is used as an auxiliary feature. The expressions of velocity v(t) and acceleration a(t) are: in, is the first-order derivative of the position vector x(t) with respect to time t. The collected data is cleaned and standardized, and the labeled data is classified by activity type to construct a training data set. Each data set contains multiple features, such as motion trajectory, speed, acceleration, and light changes. According to different activity types, the data is divided into multiple subsets, which are used as training sets and validation sets in turn. Select the SVM-CNN algorithm to generate the model according to the data characteristics and classification requirements, use the training set to train the SVM-CNN model, and finally use the validation set to validate the trained SVM-CNN model; The trained model is deployed to the system, and the sensor data is received in real time for analysis and classification. The model determines the current activity type based on the input features and outputs the classification results. The classification results are used to drive the lighting control system to achieve intelligent dimming and switch control. Through online learning and incremental learning, model parameters are continuously adjusted and optimized based on user feedback and actual operation data.

4. The intelligent lighting control system based on optical and radar perception according to claim 4 is characterized in that: The SVM-CNN model includes convolutional layers, pooling layers and fully connected layers. The convolutional layers are used to extract spatial features, the pooling layers are used for downsampling, and the fully connected layers are used to flatten and connect the outputs of the pooling layers. SVM uses the high-level features extracted by CNN for classification and adopts a one-to-many strategy in multi-classification problems.

5. The intelligent lighting control system based on optical and radar perception according to claim 4, characterized in that: When processing the data from the perception module, the processor also needs to perform person recognition on the image data captured by the optical sensor, identify each person in the room, assign a unique identifier, and set the corresponding weight according to the characteristics of the person; Set adjustment priorities based on weight and activity, and make dynamic adjustments based on the character's activity status and weight.

6. The intelligent lighting control system based on optical and radar perception according to claim 4, characterized in that: After receiving the optical information, the perception module will also collect the ambient light intensity data in real time through the optical sensor, record the time series of light changes, and when a sensor fails, the backup sensor can immediately take over its work; In each time window, the light intensity data is averaged or weighted averaged to reduce the fluctuation caused by instantaneous light changes and perform exponential weighted averaging on the light intensity data. The weight of new data points gradually decreases, and recent data has a greater impact on the results, thereby smoothing the light changes and obtaining smoothed light data. According to the smoothed light data, the brightness and switch status of the lamp are adjusted in real time, so that the brightness of the lamp can be adjusted smoothly. Use the light change buffer algorithm to smooth light changes and instantaneous light changes.

7. The intelligent lighting control system based on optical and radar perception according to claim 5, characterized in that: It is used to collect the operation data of smart lamps, analyze the operation time points, determine the frequent operation time periods and occasional operation time periods, and optimize the control strategy of the lamps based on the operation analysis results. Integrate data acquisition modules into smart lamps to record the operating status of lamps in real time; The collected operation data is sorted in chronological order to construct a time series data set. The switch time, usage duration and brightness adjustment of the lamps are analyzed, and the frequency analysis of the time series data is performed to determine the usage frequency of the lamps in different time periods of the day, and identify high-frequency usage periods and low-frequency usage periods; Perform frequency analysis on time series data to determine the frequency of lamp usage at different time periods during the day and identify high-frequency and low-frequency usage periods; Use data mining technology to analyze users' lighting usage habits in different time periods and identify high-frequency and low-frequency usage patterns; Correlate and analyze the lamp operation data with environmental parameters and user behavior data to identify the correlation between specific behaviors and lamp operation; Generate optimized lighting control strategies based on the results of operation time point analysis and operation mode recognition.

8. The intelligent lighting control system based on optical and radar perception according to claim 6, characterized in that: The lamp control prediction processing unit performs the following steps when performing the prediction control strategy: parsing the received control signal, identifying the specific operation to be performed; parsing the received control signal, identifying the specific operation to be performed; A lamp operation status model is established, taking into account the current environmental status, user behavior and historical data. The model predicts the switch status, brightness level and usage frequency of the lamp, and inputs the current environmental data, user behavior data and historical data characteristics into the model for status prediction. Finally, the model is used to calculate and predict the operating status of the lamp in the future.

9. The intelligent lighting control system based on optical and radar perception according to claim 7, characterized in that: The lamp operation analysis unit also includes the following steps when adjusting the control strategy: Monitor the current environmental status and sensor working status, environmental changes and user behavior in real time, and dynamically adjust the sensitivity of the sensor based on real-time environmental data; When the system detects a degradation in sensor performance or a drastic change in the environment, it immediately issues a warning signal and adjusts the control strategy.

Citation Information

Patent Citations

  • Fire-fighting evacuation path planning method, device and equipment based on multiple sensors

    CN114863075A

  • Indoor positioning method and system based on multi-source sensor fusion

    CN116222556A

  • Intelligent home environment sensing switch control system

    CN118131639A

Cited By

  • Airport navigation aid single lamp monitoring system based on Internet of Things communication technology

    CN119629814A