A method and system for adjusting light source state of intelligent lighting system
By acquiring and processing light intensity data in real time in an intelligent lighting system, distinguishing outliers and non-outliers data, and building a dynamic hysteresis smooth model, the problem of poor stability of light intensity data in light source state adjustment is solved, and the light source state adjustment is improved Accuracy and prediction accuracy.
Patent Information
- Application Number
- CN202410986948.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-23
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-07-23
AI Technical Summary
In the adjustment of light source state, the existing intelligent lighting system has poor time stability of the light intensity data, resulting in low accuracy of prediction results, which in turn affects the accuracy of light source state adjustment.
By obtaining the original light intensity sequence in the building in real time, determining the outlier degree at each moment, distinguishing the outlier and non-outlier data, building a dynamic hysteresis smoothing model, smoothing the light intensity, generating a stationary sequence, and then performing the light source state Adjustment.
It improves the accuracy of light source state adjustment, enhances the accuracy of light change trend prediction, overcomes the problem of poor time stability, and reduces noise and calculation complexity.
Smart Images

Figure CN118678518B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of light source control, and in particular to a method and system for adjusting the state of a light source of an intelligent lighting system. Background Art
[0002] With the development of science and technology, the lighting system in buildings is becoming more and more intelligent. The adjustment of the light source status of the intelligent lighting system helps to achieve energy saving and comfort optimization. Specifically, it can customize the lighting according to different times, different situations or different needs to meet the needs of users to the greatest extent and improve comfort. It also helps to avoid unnecessary energy waste and reduce energy consumption. In addition, the intelligent lighting system can also improve the aesthetics and use value of the building through flexible light source control.
[0003] In the prior art, the light source state of the intelligent lighting system inside the building is usually adjusted based on the time series prediction algorithm, specifically: according to the predicted trend of light changes inside the building, the light source state adjustment control algorithm is used to control the light. The existing prediction algorithm is generally based on the time stationary characteristics of the data for prediction. However, the time series changes of the light intensity inside the building are affected by many factors, such as clouds, passing vehicles and other factors that block the lighting inside the building, which makes the time stationarity of the original light intensity data poor. When the original light intensity data is used as the basic data for prediction, the accuracy of the obtained prediction results is often poor. The prior art generally uses the difference method to process data with poor time stationarity, but the order of the difference is too high, which may introduce noise, increase the complexity of the calculation and lose more information, resulting in low authenticity of the original light intensity data, so that the accuracy of the obtained prediction results is still poor, resulting in low precision of the light source state adjustment. Summary of the invention
[0004] In order to solve the technical problem that the prediction result obtained from the original light intensity data has poor accuracy, resulting in low precision in adjusting the light source state, the object of the present invention is to provide a method and system for adjusting the light source state of an intelligent lighting system. The technical solutions adopted are as follows:
[0005] An embodiment of the present invention provides a method for adjusting the state of a light source of an intelligent lighting system, the method comprising the following steps:
[0006] Real-time acquisition of the original light intensity of the light source area to be adjusted in the building at each moment within the preset period of time to form an original light intensity sequence;
[0007] According to the original light intensity at each moment in the original light intensity sequence, determine the outlier degree of the original light intensity at each moment; distinguish each outlier light intensity and each non-outlier light intensity in the original light intensity sequence according to the outlier degree;
[0008] According to each outlier light intensity in the original light intensity sequence, the size and distribution of the outlier light intensity are analyzed to determine the influence of each outlier light intensity on the overall light trend;
[0009] Determine the light intensities to be smoothed corresponding to the target non-outlier light intensities according to the outlier light intensities, the outlier light intensities, the outlier degree of the outlier light intensities, and the degree of influence on the overall light trend;
[0010] A dynamic hysteresis smoothing model is constructed, and all the light intensities to be smoothed are smoothed using the dynamic hysteresis smoothing model to obtain a steady light intensity sequence;
[0011] According to the stable light intensity sequence, the light source state of the light source area to be adjusted is adjusted.
[0012] Furthermore, determining the degree of outlier of the original illumination intensity at each moment in the original illumination intensity sequence according to the original illumination intensity at each moment includes:
[0013] The original light intensity at any moment is taken as the target light intensity, the average value of the maximum original light intensity and the minimum original light intensity in the original light intensity sequence is first determined, and then the difference between the average value and the target light intensity is calculated, and the difference is taken as the first outlier factor of the target light intensity;
[0014] Determine a second outlier factor of the target light intensity according to the original light intensity at each moment in the original light intensity sequence and its corresponding time sequence number;
[0015] The outlier degree of the target light intensity is determined according to the first outlier factor and the second outlier factor; wherein both the first outlier factor and the second outlier factor are positively correlated with the outlier degree.
[0016] Furthermore, the determining of the second outlier factor of the target light intensity according to the original light intensity at each moment in the original light intensity sequence and its corresponding time sequence number includes:
[0017] The calculation formula for the second outlier factor of the original light intensity at the tth moment is:
[0018] Where OU2 tis the second outlier factor of the original light intensity at the tth moment, T is the number of original light intensities at all moments in the original light intensity sequence, t and t' are the time sequence numbers of the original light intensities in the original light intensity sequence, e is a natural constant, S t is the original light intensity at the tth moment, S t' is the original light intensity at the t'th moment, || is the absolute value function.
[0019] Furthermore, the step of analyzing the size and distribution of the outlier illumination intensities in the original illumination intensity sequence to determine the influence of the outlier illumination intensities on the overall illumination trend includes:
[0020] Take any outlier light intensity as the target outlier light intensity, count the number of other outlier light intensities that are located around the target outlier light intensity and are continuously adjacent to it, and record it as the adjacent number; take all other outlier light intensities in the original light intensity sequence except the target outlier light intensity as the comparison data of the target outlier light intensity;
[0021] The influence of the target outlier light intensity on the overall light trend is determined according to the target outlier light intensity and the adjacent number of the target outlier light intensity, the adjacent number of each comparison data of the target outlier light intensity, and the other outlier light intensities that are located around the target outlier light intensity and are continuously adjacent.
[0022] Further, determining the influence of the target outlier light intensity on the overall light trend according to the target outlier light intensity and the number of adjacent target outlier light intensity, the number of adjacent comparison data of the target outlier light intensity, and other outlier light intensities located around the target outlier light intensity and continuously adjacent to it, includes:
[0023] Calculate the average value of the ratio between the adjacent numbers of the target outlier light intensity and the adjacent numbers of its respective comparison data as the first influencing factor;
[0024] Calculate the average of the differences between the target outlier light intensity and each of the other outlier light intensities that are located around the target outlier light intensity and are continuously adjacent to the target outlier light intensity as the second influencing factor;
[0025] The influence degree of the target outlier light intensity on the overall light trend is determined according to the first influencing factor and the second influencing factor; wherein the first influencing factor and the second influencing factor are both positively correlated with the influence degree.
[0026] Further, the step of determining the illumination intensities to be smoothed corresponding to the target non-outlier illumination intensities according to the non-outlier illumination intensities, the outlier illumination intensities, the outlier degree of the outlier illumination intensities, and the degree of influence on the overall illumination trend of the outlier illumination intensities includes:
[0027] In the original light intensity sequence, the non-outlier light intensity that is before and adjacent to the outlier light intensity is taken as the target non-outlier light intensity, thereby obtaining each target non-outlier light intensity;
[0028] The formula for calculating the number of lags of the vth target non-outlier light intensity is:
[0029] Where, SN v is the number of lags of the non-outlier light intensity of the vth target, m v is the number of other outlier light intensities that are located around the vth target non-outlier light intensity and are continuously adjacent to it, Norm is a linear normalization function, S v is the non-outlier light intensity of the vth target, is the average value of the outlier light intensities that are located around and adjacent to the vth target non-outlier light intensity, g is the serial number of the other outlier light intensities that are located around and adjacent to the vth target non-outlier light intensity, OU v,g K is the outlier degree of the g-th outlier light intensity located around and adjacent to the v-th target non-outlier light intensity, v,g is the influence of the g-th outlier light intensity located around and adjacent to the v-th target non-outlier light intensity on the overall light trend, || is the absolute value function;
[0030] In the original light intensity sequence, the number of lagged original light intensities behind the target non-outlier light intensity is used as the light intensities to be smoothed of the target non-outlier light intensity.
[0031] Furthermore, the expression calculation formula of the dynamic hysteresis smoothing model is:
[0032] Where SG is the dynamic hysteresis smoothing model function, SG(SN v ) is the lag number of the vth target non-outlier light intensity, the smoothed value of the light intensity to be smoothed, e is a natural constant, S v is the non-outlier light intensity of the vth target, is the average value of the outlier light intensities that are located around the vth target non-outlier light intensity and are continuously adjacent to it, and || is the absolute value function.
[0033] Furthermore, adjusting the light source state of the light source area to be adjusted according to the stable light intensity sequence includes:
[0034] First, the steady light intensity sequence is processed using a time series prediction algorithm to obtain a predicted light intensity sequence. Then, based on the predicted light intensity sequence, a machine learning algorithm is used to adjust the light source state of the light source area to be adjusted.
[0035] Further, the step of distinguishing each outlier light intensity and each non-outlier light intensity in the original light intensity sequence according to the outlier degree includes:
[0036] An outlier degree threshold is set, and the original light intensity in the original light intensity sequence that is greater than the outlier degree threshold is taken as the outlier light intensity, and the original light intensity in the original light intensity sequence that is not greater than the outlier degree threshold is taken as the non-outlier light intensity.
[0037] An embodiment of the present invention further provides a light source state adjustment system for an intelligent lighting system, comprising a processor and a memory, wherein the processor is used to process instructions stored in the memory to implement a light source state adjustment method for the intelligent lighting system.
[0038] The present invention has the following beneficial effects:
[0039] In order to improve the accuracy of dynamic adjustment of light sources, the present invention provides a method and system for adjusting the state of light sources of an intelligent lighting system, which mainly relates to the field of light source control technology, and is mainly used in the field of dynamic adjustment of light source states. First, the original light intensity sequence is obtained to determine the degree of outliers in the original light intensity. The degree of outliers can not only be used to simply distinguish outlier data and non-outlier data in the sequence, but also provide data support for the subsequent determination of the light intensity to be smoothed; the data in the sequence is preliminarily divided into outlier data and non-outlier data based on the degree of outliers, which helps to simply screen out outlier data that has a greater impact on temporal stability, so as to facilitate subsequent data analysis, and screening out non-outlier data can provide basic data for smoothing processing, thereby improving the reliability of corrected smoothing processing; secondly, the size and distribution of the outlier light intensity are used to determine the intensity of each outlier. The influence of intensity on the overall lighting trend can quantify the complexity of the generation of each outlier lighting intensity and the duration of the influence, which is convenient for the subsequent determination of each light intensity to be smoothed corresponding to each target non-outlier lighting intensity; then, when determining each light intensity to be smoothed, not only the outlier data initially screened out is taken into account, but also the non-outlier data after the outlier data is taken into account, which improves the coverage of the smoothing process, and then helps to improve the temporal stability of the entire original lighting intensity sequence and enhance the accuracy of subsequent prediction results; finally, based on the more accurate lighting change trend prediction results, the accuracy of light source state adjustment can be further improved. BRIEF DESCRIPTION OF THE DRAWINGS
[0040] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0041] Figure 1 This is a flow chart of a method for adjusting the state of a light source of an intelligent lighting system according to an embodiment of the present invention;
[0042] Figure 2 The flowchart is a step of determining the outlier degree of target light intensity in an embodiment of the present invention. DETAILED DESCRIPTION
[0043] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation methods, structures, features and effects of the technical solutions proposed by the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.
[0044] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
[0045] The embodiment of the present invention provides a method for adjusting the light source state of an intelligent lighting system. Figure 1 As shown, the following steps are included:
[0046] S1, real-time acquisition of the original light intensity of the light source area to be adjusted in the building at each moment within a preset period of time to form an original light intensity sequence.
[0047] The main purpose of this embodiment is to intelligently adjust the light source status of the light source area to be adjusted in the building. In order to reduce unnecessary energy consumption and improve lighting comfort, before intelligently adjusting the light source status of the light source area to be adjusted, it is first necessary to obtain the lighting information of the natural light in the light source area to be adjusted.
[0048] In the existing lighting systems inside buildings, in order to adjust the light source status inside the building according to the lighting conditions of natural light, the intelligent lighting system generally arranges optical sensors that collect natural light information in each light source area to be adjusted, so as to collect the light intensity information of natural light in the light source area to be adjusted in real time, and then realize the light source status adjustment.
[0049] The implementation process of adjusting the light source state of each light source area to be adjusted is the same. For the convenience of description, this embodiment takes any light source area to be adjusted as an example to adjust the light source state of the intelligent lighting system. Therefore, an optical sensor is used to collect the light intensity of natural light at each moment in the current hour in a light source area to be adjusted, and the light intensity of natural light is recorded as the original light intensity to form an original light intensity sequence.
[0050] In addition, after obtaining the original light intensity sequence, in order to improve the accuracy of the data to a certain extent, the data in the original light intensity sequence can be subjected to median filtering to obtain a data sequence after filtering, and the original light intensity sequence is updated to a data sequence after filtering. The implementation process of median filtering is a prior art and is not within the scope of protection of the present invention. It will not be elaborated in detail here. Of course, the implementer can also use other filtering methods to implement the filtering of the original light intensity sequence, which is not specifically limited here.
[0051] In this embodiment, the preset period for collecting light intensity can be set to 1 hour, and the time interval between two adjacent collection moments can be set to 10 seconds, that is, the collection frequency is once every 10 seconds. The collection period and collection frequency of data can be set by the implementer according to the specific actual situation, and are not specifically limited here.
[0052] So far, this embodiment has obtained the illumination intensity sequence of natural light in the light source area to be adjusted within 1 hour.
[0053] It should be noted that the method for adjusting the light source state of the light source area to be adjusted in this embodiment is to use the real-time light data to predict the light change trend of the light source area to be adjusted, and perform light regulation according to the light change trend. However, the existing prediction algorithms are generally based on the time stationary characteristics of the data for prediction, but the time series changes of the illumination intensity data of natural light are affected by many factors, which makes the time stationarity of the original illumination intensity data poor. When the original illumination intensity data is used for prediction, the accuracy of the obtained prediction results is often poor. The existing technology generally uses the difference method to process data with poor time stationarity. After processing using the difference method, if the order of the difference is too high, the prediction result may still be inaccurate.
[0054] In order to enable the intelligent lighting system in the light source area to be adjusted to achieve more accurate light source adjustment and avoid frequent switching of light source state adjustments with large differences, this embodiment uses a dynamic hysteresis smoothing method to process the collected original light intensity sequence, which overcomes the defect of poor time stability to a certain extent. Specifically, a dynamic hysteresis smoothing model is established, and the dynamic hysteresis smoothing model is used to smooth all the light intensity data to be smoothed, so as to obtain the smoothed data. The detailed steps are steps S2 to S5.
[0055] S2, determining the outlier degree of the original light intensity at each moment in the original light intensity sequence according to the original light intensity at each moment; and distinguishing each outlier light intensity and each non-outlier light intensity in the original light intensity sequence according to the outlier degree.
[0056] Affected by factors such as the external environment, there are some outlier data in the original light intensity sequence. For example, clouds, passing vehicles, or building shadows block the lighting channels of the light source area to be adjusted, resulting in outlier data in the original light intensity sequence. The greater the impact of outlier data on the temporal stationarity of the original light intensity sequence, the greater the possibility of lagging smoothing of outlier data. Therefore, it is necessary to filter out outlier data from the original light intensity sequence. First, it is necessary to determine the outlier degree of the original light intensity at each moment in the original light intensity sequence. The outlier degree can be used to quantify the possibility that the original light intensity is an outlier. The greater the outlier degree, the greater the possibility that the original light intensity is an outlier.
[0057] In the first step, according to the original light intensity at each moment in the original light intensity sequence, the outlier degree of the original light intensity at each moment is determined.
[0058] In this embodiment, the calculation process of the outlier degree of each original light intensity is consistent. In order to reduce unnecessary descriptions and make it easier to understand the solution, any original light intensity is used as an example to determine the outlier degree, and any original light intensity is used as the target light intensity, such as Figure 2 As shown, the specific steps of determining the outlier degree of the target light intensity may include:
[0059] In the first sub-step, the average value of the maximum original light intensity and the minimum original light intensity in the original light intensity sequence is first determined, and then the difference between the average value and the target light intensity is calculated, and the difference is used as the first outlier factor of the target light intensity.
[0060] In the light source area to be adjusted of the building, when there is no influence of unexpected factors, the overall original light intensity presents a relatively stable linear change over time, that is, when there is no influence of unexpected factors, the light intensity of natural light is only affected by time; when there is an influence of unexpected factors, the overall original light intensity will change significantly, so this embodiment uses the median of the original light intensity sequence to represent the lighting trend of the light source area to be adjusted without the influence of unexpected factors, and then uses the numerical difference between the target light intensity and the median to quantify the degree of outlier of the target light intensity.
[0061] Among them, the first outlier factor can characterize the overall centrifugal trend of the target light intensity. The larger the overall centrifugal trend, the less the target light intensity meets the overall light intensity change trend of the sequence. The greater the outlier degree of the target light intensity, the greater the possibility that the target light intensity is an outlier data.
[0062] In the second sub-step, according to the original illumination intensity at each moment in the original illumination intensity sequence and its corresponding time sequence number, a second outlier factor of the target illumination intensity is determined.
[0063] When determining the second outlier factor, the target light intensity is taken as the core, and the timing differences and numerical differences between the target light intensity and other original light intensities except the target light intensity are analyzed. The larger the difference, the greater the difference between the target light intensity and the original light intensity at a closer distance. Since the change in light intensity is a continuous and slow process, the greater the degree of outlier of the target light intensity at this time, the greater the possibility that the target light intensity is an outlier data.
[0064] In this embodiment, the target illumination intensity may be the original illumination intensity at the t-th moment in the original illumination intensity sequence, and the calculation formula of the second outlier factor of the original illumination intensity at the t-th moment may be:
[0065] Where OU2 t is the second outlier factor of the original light intensity at the tth moment, T is the number of original light intensities at all moments in the original light intensity sequence, t and t' are the time sequence numbers of the original light intensities in the original light intensity sequence, e is a natural constant, S t is the original light intensity at the tth moment, S t' is the original light intensity at the t'th moment, || is the absolute value function.
[0066] In the calculation formula of the second outlier factor, e-|tt′| can be expressed as |S t -S t' |The weight of |, the larger the distance |t-t'| between the tth moment and the t'th moment, |S t -St' The smaller the weight ratio of | is; the second outlier factor can characterize the time series impact difference value of the target light intensity. The larger the second outlier factor is, the greater the difference between the original light intensity at the tth moment and the original light intensities at multiple moments with a closer distance in the sequence is, the greater the outlier degree of the original light intensity at the tth moment is, and the greater the possibility that the original light intensity at the tth moment is an outlier data.
[0067] In the third sub-step, the outlier degree of the target light intensity is determined according to the first outlier factor and the second outlier factor.
[0068] In this embodiment, the first outlier factor and the second outlier factor are both positively correlated with the outlier degree. The first outlier factor and the second outlier factor can be multiplied, added or calculated in other ways to determine the outlier degree. The positive correlation means that the outlier degree increases with the increase of the first outlier factor and the second outlier factor, or decreases with the decrease of the first outlier factor and the second outlier factor.
[0069] As an example, the calculation formula for the outlier degree of the original light intensity at the t-th moment can be:
[0070] In the formula, OU t is the outlier degree of the original light intensity at the tth moment, maxS is the maximum original light intensity in the original light intensity sequence, minS is the minimum original light intensity in the original light intensity sequence, is the median of the original light intensity sequence, S t is the original light intensity at the tth moment, || is the absolute value function, Norm is the linear normalization function, is the first outlier factor of the original light intensity at the tth moment, OU2 t is the second outlier factor of the original light intensity at the t-th moment.
[0071] In the second step, each outlier light intensity and each non-outlier light intensity in the original light intensity sequence are distinguished according to the degree of outlier.
[0072] The degree of outliers indicates the difference in the changing trend of the original light intensity at each moment compared with the data with a relatively gentle change. The outlier data can be judged according to the degree of outliers of the original light intensity at each moment. The judgment method adopted in this embodiment is threshold judgment. The specific judgment process may include:
[0073] In this embodiment, an outlier degree threshold is set, and the original light intensity in the original light intensity sequence that is greater than the outlier degree threshold is used as the outlier light intensity, and the original light intensity in the original light intensity sequence that is not greater than the outlier degree threshold is used as the non-outlier light intensity.
[0074] Among them, the outlier light intensity is the outlier data in the original light intensity sequence, and the non-outlier light intensity is the original light intensity other than the outlier data in the original light intensity sequence; the outlier degree threshold is set to 0.5, and the implementer can set the outlier degree threshold size according to the actual situation, and no specific limitation is made here.
[0075] So far, this embodiment divides all the original light intensities in the original light intensity sequence into outlier data and non-outlier data according to the outlier degree.
[0076] S3, analyzing the size and distribution of the outlier light intensity according to each outlier light intensity in the original light intensity sequence, and determining the influence of each outlier light intensity on the overall light trend.
[0077] First of all, it should be noted that the outlier light intensity is more non-temporally stationary than the original light intensity data, but a single outlier light data has little effect on the overall prediction. Therefore, the influence of outlier data on the overall light trend can be quantitatively analyzed based on the persistence of the outlier light intensity and the degree of gap with the non-outlier data, so as to provide data conditions for the subsequent determination of the light intensity data to be smoothed.
[0078] The calculation process of the influence of each outlier light intensity on the overall light trend is consistent. In order to avoid repeated description of the same technical features, any outlier light intensity is taken as an example to determine the influence on the overall light trend, and any outlier light intensity is taken as the target outlier light intensity. The specific implementation steps may include:
[0079] In the first step, the number of other outlier light intensities that are located around the target outlier light intensity and are continuously adjacent to it is counted and recorded as the adjacent number; the other outlier light intensities in the original light intensity sequence except the target outlier light intensity are used as the comparison data of the target outlier light intensity.
[0080] In this embodiment, the adjacent number refers to the number of original light intensities selected around the target outlier light intensity, which are adjacent to the target outlier light intensity and are continuous; the adjacent number of the target outlier light intensity is counted to analyze the duration of the outlier of the light intensity. The size of the adjacent number is related to the acquisition frequency of the original light intensity, so the size of the adjacent number is not a fixed unique value. Generally, there is no case where the adjacent number is 0. For the convenience of description, the outlier light intensities other than the target outlier light intensity used for the subsequent calculation of the degree of influence are collectively referred to as the comparison data of the target outlier light intensity. At the same time, the cluster of the target outlier light intensity and other outlier light intensities around it and continuously adjacent to it can also be referred to as the outlier light intensity cluster of the target outlier light intensity.
[0081] In the second step, the influence of the target outlier light intensity on the overall light trend is determined based on the target outlier light intensity and the adjacent number of the target outlier light intensity, the adjacent number of each comparison data of the target outlier light intensity, and the other outlier light intensities that are located around the target outlier light intensity and are continuously adjacent.
[0082] When analyzing the influence of the target outlier light intensity on the overall light trend, two aspects are mainly used for calculation. The first aspect is the continuous influence of the target outlier light intensity, and the second aspect is the cluster coherence of the target outlier light intensity. The specific implementation steps may include:
[0083] In the first sub-step, the average value of the ratio between the adjacent numbers of the target outlier illumination intensity and the adjacent numbers of its respective comparison data is calculated as the first influencing factor.
[0084] In this embodiment, the light source state is adjusted by the prediction result of the continuous trend of the illumination intensity of natural light, so the influence of a single outlier illumination intensity on the overall illumination trend is small. The longer the duration of the target outlier illumination intensity, that is, the number of adjacent target outlier illumination intensity is much larger than the number of adjacent comparison data, the larger the first influencing factor is, which indicates that the influence of the target outlier illumination intensity on the overall illumination trend is greater.
[0085] In the second sub-step, an average value of the difference between the target outlier light intensity and each of the other outlier light intensities that are located around the target outlier light intensity and are continuously adjacent to the target outlier light intensity is calculated as the second influencing factor.
[0086] The second influencing factor is actually a supplement to the calculation logic of the first influencing factor. The first influencing factor can only show that there are more outliers in the target outlier illumination intensity at continuous moments, but cannot explain whether the continuous adjacent outliers are related to the target illumination intensity. Therefore, this embodiment uses the average value of the difference between the target outlier illumination intensity and other outliers located around the target outlier illumination intensity and continuously adjacent to each other to express the coherence of the target outlier illumination intensity and the outliers in the outlier illumination intensity cluster to which it belongs, and determines the second influencing factor. The smaller the second influencing factor, the greater the coherence, indicating that the target outlier illumination intensity and the outliers in the outlier illumination intensity cluster to which it belongs are more likely to be generated under the influence of the same unexpected factor, and the smaller the impact on the overall illumination intensity trend prediction; on the contrary, the larger the second influencing factor, the more different the formation causes of the target outlier illumination intensity and the other outliers in the outlier data cluster to which it belongs, the more influencing factors there are, the more complex the influencing environment of the outlier data cluster, and the greater the impact on the overall illumination intensity trend prediction.
[0087] In the third sub-step, the influence degree of the target outlier illumination intensity on the overall illumination trend is determined according to the first influencing factor and the second influencing factor.
[0088] In this embodiment, the first influencing factor and the second influencing factor are both positively correlated with the degree of influence, so the degree of influence of the target outlier illumination intensity on the overall illumination trend can be determined by a fusion method such as multiplication or addition.
[0089] As an example, the calculation formula for the influence of the nth outlier light intensity on the overall light trend can be:
[0090] In the formula, K n is the influence of the nth outlier light intensity on the overall light trend, N is the number of comparison data of the nth outlier light intensity, n' is the sequence number of the comparison data of the nth outlier light intensity, m n is the number of neighbors of the nth outlier light intensity, m n' is the number of adjacent n'th comparison data of the nth outlier light intensity, is the first influencing factor of the nth outlier light intensity, n″ is the number of other outlier light intensities located around the nth outlier light intensity and adjacent to it, S n is the nth outlier light intensity, S n” is the nth outlier light intensity that is located around the nth outlier light intensity and is continuously adjacent to it, || is the absolute value function, is the second influencing factor of the nth outlier light intensity.
[0091] So far, this embodiment obtains the influence degree of each outlier data on the overall illumination trend by analyzing the continuous influence degree of each outlier data on the overall data and the cluster correlation.
[0092] S4, determining the light intensities to be smoothed corresponding to the target non-outlier light intensities according to the non-outlier light intensities, the outlier light intensities, the outlier degrees of the outlier light intensities, and the degree of influence on the overall light trend of the outlier light intensities.
[0093] After obtaining the degree of outlier of each outlier light intensity and the degree of influence on the overall light intensity trend, in order to facilitate the subsequent smoothing of the original light intensity data that affects the overall light intensity trend prediction, improve the accuracy of subsequent light prediction and improve the accuracy of light source state adjustment, it is necessary to determine all the light intensity data to be smoothed. The specific implementation steps may include:
[0094] In the first step, in the original light intensity sequence, the non-outlier light intensity that is located before and adjacent to the outlier light intensity is used as the target non-outlier light intensity, thereby obtaining each target non-outlier light intensity.
[0095] When smoothing part of the light intensity data in the original light intensity sequence, the dynamic hysteresis smoothing model is constructed mainly based on the non-outlier light intensity as the basic data. However, since the light intensity of natural light changes over time, the non-outlier light intensity that is before and adjacent to the outlier light intensity is screened out. For ease of distinction, it is called the target non-outlier light intensity.
[0096] In the second step, the number of lags for each target non-outlier light intensity is determined.
[0097] Regarding the number of lags, there may be light intensity data lasting for a relatively short period of time in the original light intensity sequence. If the light source is adjusted based on the light intensity data, the light source state of the light source area to be adjusted will change significantly, which may easily cause instability in the change of the light source state. Therefore, this embodiment determines the number of lags of each target non-outlier light intensity and screens out light intensity data lasting for a relatively short period of time, so as to facilitate subsequent smoothing and correction processing.
[0098] In this embodiment, the calculation process of the lag number of each target non-outlier light intensity is consistent. For the convenience of description, any target non-outlier light intensity is taken as an example to determine the lag data. Any target non-outlier light intensity can be the vth target non-outlier light intensity, and v is the serial number of the target non-outlier light intensity.
[0099] As an example, the number of lags for the vth target non-outlier light intensity is calculated as:
[0100] Where, SN v is the number of lags of the non-outlier light intensity of the vth target, m v is the number of other outlier light intensities that are located around the vth target non-outlier light intensity and are continuously adjacent to it, Norm is a linear normalization function, S v is the non-outlier light intensity of the vth target, is the average value of the outlier light intensities that are located around and adjacent to the vth target non-outlier light intensity, g is the serial number of the other outlier light intensities that are located around and adjacent to the vth target non-outlier light intensity, OU v,g K is the outlier degree of the g-th outlier light intensity located around and adjacent to the v-th target non-outlier light intensity, v,g is the influence of the g-th outlier light intensity located around and adjacent to the v-th target non-outlier light intensity on the overall light trend, and || is the absolute value function.
[0101] In the calculation formula of the number of lags, the lag data refers to how many lag data need to be smoothed using the v-th target non-outlier light intensity; the greater the difference between the v-th target non-outlier light intensity and the outlier light intensity after the v-th target non-outlier light intensity, the higher the outlier degree of each outlier light intensity located around and adjacent to the v-th target non-outlier light intensity, and the greater the impact on the overall light intensity trend, the larger the range of using the v-th target non-outlier light intensity data as smoothing data; the greater the difference between the v-th target non-outlier light intensity and the outlier light intensity after the v-th target non-outlier light intensity, the less accurate the result of the outlier light intensity after the v-th target non-outlier light intensity for light prediction, and the greater the outlier degree and the impact on the overall light trend prediction, which means that the natural light in the light source area to be adjusted is affected by more external factors and the greater the degree of influence, and the more the data after the v-th target non-outlier light intensity needs to be smoothed.
[0102] In the third step, in the original light intensity sequence, the lagged original light intensities after the target non-outlier light intensity are used as the light intensities to be smoothed of the target non-outlier light intensity.
[0103] It should be noted that the data properties of the light intensity data to be smoothed include all outlier light intensity data and part of the non-outlier light intensity. A part of the non-outlier light intensity is selected to compensate for the possible errors in distinguishing outliers and non-outliers to a certain extent. The light intensity to be smoothed determined by combining the outlier degree and influence degree of the outlier data improves the coverage of the smoothing processing and further improves the accuracy of the dynamic smoothing processing of the original light intensity sequence.
[0104] So far, this embodiment has screened out all the light intensity data to be smoothed from the original light intensity sequence.
[0105] S5, constructing a dynamic hysteresis smoothing model, and using the dynamic hysteresis smoothing model to smooth all the light intensities to be smoothed to obtain a steady light intensity sequence.
[0106] The first step is to construct a dynamic lag smoothing model.
[0107] The target non-outlier light intensity is the light intensity data that changes normally with time. The outlier changes in the lagged number of light intensities to be smoothed after the target non-outlier light intensity are accidental, but also contain some trend change characteristics of normal light. Therefore, when constructing a dynamic lag smoothing model, it is necessary to make the weight of the target non-outlier light intensity larger to ensure the temporal stability of the data. At the same time, a certain weight is also given to the outlier data after the target non-outlier light intensity, so that on the basis of ensuring the temporal stability of the data, some outlier data characteristics are retained.
[0108] In this embodiment, the expression calculation formula of the dynamic hysteresis smoothing model can be:
[0109] Where SG is the dynamic hysteresis smoothing model function, SG(SN v ) is the lag number of the vth target non-outlier light intensity, the smoothed value of the light intensity to be smoothed, e is a natural constant, S v is the non-outlier light intensity of the vth target, is the average value of the outlier light intensities that are located around the vth target non-outlier light intensity and are continuously adjacent to it, and || is the absolute value function.
[0110] It should be noted that the calculation logic of the dynamic lag smoothing model is a weighted average algorithm, which can be used to calculate the smoothed value of the lagged number of light intensities to be smoothed of each target non-outlier light intensity; the greater the difference between the vth target non-outlier light intensity and the average value of each surrounding and continuously adjacent outlier light intensity, the greater the weight of the vth target non-outlier light intensity, and vice versa.
[0111] In the second step, all the light intensities to be smoothed are smoothed using a dynamic hysteresis smoothing model to obtain a steady light intensity sequence.
[0112] For each target non-outlier light intensity to be smoothed, the light intensity to be smoothed is processed using a dynamic hysteresis smoothing model to obtain a smoothed original light intensity sequence, and the smoothed original light intensity sequence is used as a stationary light intensity sequence. Among them, the smoothed values of each light intensity to be smoothed of the same target non-outlier light intensity are the same.
[0113] So far, this embodiment has obtained a stable light intensity sequence with good temporal stability.
[0114] S6, adjusting the light source state of the light source area to be adjusted according to the stable light intensity sequence.
[0115] In this embodiment, the steady light intensity sequence is first processed using a time series prediction algorithm to obtain a predicted light intensity sequence; then, based on the predicted light intensity sequence, a machine learning algorithm is used to adjust the light source state of the light source area to be adjusted.
[0116] Among them, the predicted light intensity sequence can characterize the trend of light change, the time series prediction algorithm adopted can be ARIMA (Auto regressive Integrated Moving Average model), and the machine learning algorithm can be an optimization algorithm, such as a genetic algorithm, a simulated annealing algorithm, etc. The implementation process of ARIMA prediction and optimization algorithm adjustment are both existing technologies, which are not within the protection scope of the present invention and will not be elaborated in detail here.
[0117] An embodiment of the present invention further provides a light source state adjustment system for an intelligent lighting system, comprising a processor and a memory, wherein the processor is used to process instructions stored in the memory to implement a light source state adjustment method for the intelligent lighting system.
[0118] To summarize, in order to be able to more accurately adjust the light source state of the area to be adjusted in the building, this embodiment proposes a dynamic hysteresis smoothing method based on light changes, which is used to process the data that needs to be corrected and smoothed in the original light intensity sequence, thereby improving the time stability of the collected original light intensity sequence, overcoming the influence of external environmental factors when adjusting the light source to a certain extent, and further improving the accuracy of light source adjustment in the intelligent lighting system.
[0119] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.
Claims
1. A method for adjusting the light source state of an intelligent lighting system, characterized in that: The following steps are involved: Real-time acquisition of the original light intensity of the light source area to be adjusted in the building at each moment within the preset period of time to form an original light intensity sequence; According to the original light intensity at each moment in the original light intensity sequence, determine the outlier degree of the original light intensity at each moment; Distinguish each outlier light intensity and each non-outlier light intensity in the original light intensity sequence according to the outlier degree; According to each outlier light intensity in the original light intensity sequence, the size and distribution of the outlier light intensity are analyzed to determine the influence of each outlier light intensity on the overall light trend; Determine the light intensities to be smoothed corresponding to the target non-outlier light intensities according to the outlier light intensities, the outlier light intensities, the outlier degree of the outlier light intensities, and the degree of influence on the overall light trend; A dynamic hysteresis smoothing model is constructed, and all the light intensities to be smoothed are smoothed using the dynamic hysteresis smoothing model to obtain a steady light intensity sequence; According to the stable light intensity sequence, the light source state of the light source area to be adjusted is adjusted; The step of analyzing the size and distribution of the outlier light intensity in the original light intensity sequence and determining the influence of each outlier light intensity on the overall light trend includes: Take any outlier light intensity as the target outlier light intensity, count the number of other outlier light intensities that are located around the target outlier light intensity and are continuously adjacent to it, and record it as the adjacent number; take all other outlier light intensities in the original light intensity sequence except the target outlier light intensity as the comparison data of the target outlier light intensity; According to the target outlier light intensity and the adjacent number of the target outlier light intensity, the adjacent number of each comparison data of the target outlier light intensity, and each other outlier light intensity located around the target outlier light intensity and continuously adjacent, determine the influence of the target outlier light intensity on the overall light trend; The step of determining the influence of the target outlier light intensity on the overall light trend according to the target outlier light intensity and the number of adjacent target outlier light intensity, the number of adjacent comparison data of the target outlier light intensity, and other outlier light intensities located around the target outlier light intensity and continuously adjacent to the target outlier light intensity comprises: Calculate the average value of the ratio between the adjacent numbers of the target outlier light intensity and the adjacent numbers of its respective comparison data as the first influencing factor; Calculate the average of the differences between the target outlier light intensity and each of the other outlier light intensities that are located around the target outlier light intensity and are continuously adjacent to the target outlier light intensity as the second influencing factor; Determine the influence degree of the target outlier light intensity on the overall light trend according to the first influencing factor and the second influencing factor; wherein the first influencing factor and the second influencing factor are both positively correlated with the influence degree; The step of determining the illumination intensities to be smoothed corresponding to the target non-outlier illumination intensities according to the non-outlier illumination intensities, the outlier illumination intensities, the outlier degree of the outlier illumination intensities, and the degree of influence on the overall illumination trend of the outlier illumination intensities includes: In the original light intensity sequence, the non-outlier light intensity that is before and adjacent to the outlier light intensity is taken as the target non-outlier light intensity, thereby obtaining each target non-outlier light intensity; The formula for calculating the number of lags of the vth target non-outlier light intensity is: Where, SN v is the number of lags of the non-outlier light intensity of the vth target, m v is the number of other outlier light intensities that are located around the vth target non-outlier light intensity and are continuously adjacent to it, Norm is a linear normalization function, S v is the non-outlier light intensity of the vth target, is the average value of the outlier light intensities that are located around and adjacent to the vth target non-outlier light intensity, g is the serial number of the other outlier light intensities that are located around and adjacent to the vth target non-outlier light intensity, OU v,g K is the outlier degree of the g-th outlier light intensity located around and adjacent to the v-th target non-outlier light intensity, v,g is the influence of the g-th outlier light intensity located around and adjacent to the v-th target non-outlier light intensity on the overall light trend, | | is the absolute value function; In the original light intensity sequence, the lagged original light intensities after the target non-outlier light intensity are used as the light intensities to be smoothed of the target non-outlier light intensity; The expression calculation formula of the dynamic hysteresis smoothing model is: Where SG is the dynamic hysteresis smoothing model function, SG(SN v ) is the lag number of the vth target non-outlier light intensity, the smoothed value of the light intensity to be smoothed, e is a natural constant, S v is the non-outlier light intensity of the vth target, is the average value of the outlier light intensities that are located around the vth target non-outlier light intensity and are continuously adjacent to it, and | | is the absolute value function.
2. The method for adjusting the light source state of an intelligent lighting system according to claim 1, characterized in that: The step of determining the outlier degree of the original illumination intensity at each moment according to the original illumination intensity at each moment in the original illumination intensity sequence includes: The original light intensity at any moment is taken as the target light intensity, the average value of the maximum original light intensity and the minimum original light intensity in the original light intensity sequence is first determined, and then the difference between the average value and the target light intensity is calculated, and the difference is taken as the first outlier factor of the target light intensity; Determine a second outlier factor of the target light intensity according to the original light intensity at each moment in the original light intensity sequence and its corresponding time sequence number; The outlier degree of the target light intensity is determined according to the first outlier factor and the second outlier factor; wherein both the first outlier factor and the second outlier factor are positively correlated with the outlier degree.
3. The method for adjusting the light source state of an intelligent lighting system according to claim 2, characterized in that: The step of determining the second outlier factor of the target light intensity according to the original light intensity at each moment in the original light intensity sequence and its corresponding time sequence number comprises: The calculation formula for the second outlier factor of the original light intensity at the tth moment is: Where OU2 t is the second outlier factor of the original light intensity at the tth moment, T is the number of original light intensities at all moments in the original light intensity sequence, t and t' are the time sequence numbers of the original light intensities in the original light intensity sequence, e is a natural constant, S t is the original light intensity at the tth moment, S t' is the original light intensity at the t'th moment, | | is the absolute value function.
4. The method for adjusting the light source state of an intelligent lighting system according to claim 1, characterized in that: The step of adjusting the light source state of the light source area to be adjusted according to the stable light intensity sequence includes: First, the steady light intensity sequence is processed using a time series prediction algorithm to obtain a predicted light intensity sequence. Then, based on the predicted light intensity sequence, a machine learning algorithm is used to adjust the light source state of the light source area to be adjusted.
5. The method for adjusting the light source state of an intelligent lighting system according to claim 1, characterized in that: The step of distinguishing each outlier light intensity and each non-outlier light intensity in the original light intensity sequence according to the outlier degree includes: An outlier degree threshold is set, and the original light intensity in the original light intensity sequence that is greater than the outlier degree threshold is taken as the outlier light intensity, and the original light intensity in the original light intensity sequence that is not greater than the outlier degree threshold is taken as the non-outlier light intensity.
6. A light source state adjustment system for an intelligent lighting system, characterized in that: The invention comprises a processor and a memory, wherein the processor is used to process instructions stored in the memory to implement a method for adjusting the state of a light source of an intelligent lighting system as claimed in any one of claims 1 to 5.
