CN115696690B - Distributed intelligent building illumination self-adaptive energy-saving control method - Google Patents
Distributed intelligent building illumination self-adaptive energy-saving control method Download PDFInfo
- Publication number
- CN115696690B CN115696690B CN202211588933.8A CN202211588933A CN115696690B CN 115696690 B CN115696690 B CN 115696690B CN 202211588933 A CN202211588933 A CN 202211588933A CN 115696690 B CN115696690 B CN 115696690B
- Authority
- CN
- China
- Prior art keywords
- illumination
- sensor
- category
- energy
- sequence
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02B—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO BUILDINGS, e.g. HOUSING, HOUSE APPLIANCES OR RELATED END-USER APPLICATIONS
- Y02B20/00—Energy efficient lighting technologies, e.g. halogen lamps or gas discharge lamps
- Y02B20/40—Control techniques providing energy savings, e.g. smart controller or presence detection
Abstract
The invention relates to the technical field of illumination control, and provides a distributed intelligent building illumination self-adaptive energy-saving control method, which comprises the following steps: obtaining historical illumination data, sensor and energy-saving lamp information in a building; acquiring a first illumination change sequence, analyzing a first distance expression between adjacent sequences to acquire first segmentation points of all days, performing superposition analysis to acquire a plurality of first categories, and analyzing and dividing category credibility to acquire a plurality of second categories; performing weighted summation and cosine similarity analysis on the first illumination change sequence to obtain a third illumination change sequence; obtaining a plurality of third categories by division, obtaining a heteromorphic graph according to illumination data in the third categories and information of the sensor and the energy-saving lamp, and obtaining a second control parameter through a neural network; and controls energy saving lamps in the building. The invention aims to solve the problem that illumination self-adaptive adjustment cannot be processed in a module-distributed mode.
Description
Technical Field
The invention relates to the field of illumination control, in particular to a distributed intelligent building illumination self-adaptive energy-saving control method.
Background
Traditional wisdom building illumination control mode is through the change of ambient light come to throw light on to the room lamp and control the regulation adaptively. However, some time changes of the ambient light are large, so that the illumination self-adaption changes frequently, and the internal circuit is easy to be unstable, so that potential safety hazards are caused; meanwhile, if the same treatment is adopted for the whole lighting in the building according to the change of the ambient light, the energy consumption is increased, and the self-adaptation of different areas cannot be realized. Therefore, there is a need for a distributed adaptive lighting control method for processing different areas in a building, which groups areas with similar lighting changes, adaptively controls lighting devices in the areas included in different groups according to lighting conditions of different time periods in each group, implements distributed grouping processing, and reduces power consumption while maintaining circuit stability.
Disclosure of Invention
The invention provides a distributed intelligent building illumination self-adaptive energy-saving control method, which aims to solve the problem that the existing illumination self-adaptive adjustment cannot be processed in a distributed mode according to modules, and adopts the following technical scheme:
one embodiment of the invention provides a distributed intelligent building illumination self-adaptive energy-saving control method, which comprises the following steps:
acquiring first position information of an illumination sensor inside a building and second position information of an energy-saving lamp;
acquiring a first illumination change sequence of each sensor every day, acquiring a first distance between the first illumination change sequences of adjacent sensors and the occurrence frequency of each first distance, acquiring first division points of all the first illumination change sequences every day according to the occurrence frequency of all the first distances, superposing the first division points of all the days, acquiring a second distance of each first division point, acquiring a plurality of first categories by multi-threshold division of the second distances, acquiring category credibility according to the second distance of the first division points in each first category, comparing the category credibility with a first preset threshold to acquire a plurality of second division points, and dividing all the sensors according to the second division points to acquire a plurality of second categories;
acquiring a first weight value of each first illumination change sequence according to the first division point and the second division point, weighting illumination data of the first illumination change sequence of each sensor according to the first weight value to obtain a second illumination change sequence of each sensor, and acquiring a third illumination change sequence of each second category according to cosine similarity between the second illumination change sequences of each sensor in each second category;
dividing each third illumination change sequence by multi-threshold segmentation to obtain a plurality of third categories, obtaining a compensation light map structure according to illumination data in the third categories and first position information, obtaining an energy-saving lamp map structure according to second position information, connecting the compensation light map structure and the energy-saving lamp map structure according to the first position information and the second position information to form an abnormal composition, and obtaining a second control parameter of the third categories according to the abnormal composition;
and controlling the energy-saving lamp in the building according to the second control parameter.
Optionally, the first illumination change sequence of each sensor every day represents illumination data obtained by dividing each day as a period in historical illumination data collected by each sensor, the illumination data of each day of each sensor includes illumination data collected by a plurality of time nodes, and the illumination data of the plurality of time nodes form an illumination data sequence, which is the first illumination change sequence of each sensor every day.
Optionally, the obtaining of the first segmentation points of all the first illumination change sequences every day according to the first distance greater than the segmentation threshold obtained by the threshold segmentation includes the specific method that:
according to the frequency of the first distance, obtaining a division threshold of the first distance through an OTSU threshold segmentation algorithm to obtain a plurality of first distances larger than the division threshold, dividing the sensor sequence according to the first distances, wherein the corresponding segmentation points are first segmentation points of all first illumination change sequences every day, and the first segmentation points divide the sensors every day into a plurality of initial categories;
the sensor sequence is to acquire the sensor according to the first position information.
Optionally, the obtaining of the category credibility includes a specific method that:
wherein the content of the first and second substances,
/>
denotes the fourth QUOTE->
/>
A class confidence of a first class>
/>
Indicates a number of days in the historical lighting data, and->
/>
Indicates the fifth->
/>
Number of elements in a first category>
/>
Indicates the fifth->
/>
An element variance value in a first category, based on a value of a square or a square>
/>
Expressing an exponential function with a natural constant as a base; the elements in the first category are second distances of the first segmentation points on different days.
Optionally, the specific obtaining method of the second illumination change sequence of each sensor is as follows:
wherein the content of the first and second substances,
/>
indicates the ^ th or ^ th lighting change sequence of a certain sensor>
/>
Spot illumination data, <' > based on the spot illumination data>
/>
Indicates a number of days in the historical lighting data, and->
/>
Indicates that the sensor is up to->
/>
The fifth ^ in the first sequence of illumination changes of the day>
/>
The illumination data of the point is taken>
/>
Indicates that the sensor is up to->
/>
A first weight value of a first sequence of lighting changes of a day.
Optionally, the specific obtaining method of the first weight value of the first illumination variation sequence is as follows:
the first weight value is obtained from the similarity mean value of left and right first segmentation points of an initial category to which a certain first illumination change sequence belongs and left and right second segmentation points of a second category to which the first illumination change sequence belongs.
Optionally, the obtaining of the compensation light pattern structure includes the specific method:
taking the illumination sensor in any one second category and illumination data of different third categories in the second category as node information, taking the distance acquired by the first position information as a side of a graph to form an environment light graph structure, and replacing the difference value between the illumination data in the node information and the saturation light intensity to acquire a compensation light graph structure; the saturated light intensity represents a light intensity that makes human visual perception good.
Compared with the existing control mode, the invention at least has the following beneficial effects: historical illumination data of different areas in the building are collected, representative illumination data of different areas every day are obtained through weighted summation of data of multiple days, and the illumination data can better represent illumination conditions of different areas every day; the illumination areas which are close to each other along with the change of time are divided into the same group for grouping control, so that frequent parameter change is avoided, and the stability of the circuit is maintained; meanwhile, control parameters under different working modes are calculated through the neural network with the heterogeneous graph structure, the optimal energy-saving control effect is favorably obtained, and the loss of electric energy is reduced.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the description of the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a schematic flow chart of a distributed intelligent building lighting adaptive energy-saving control method according to an embodiment of the present invention;
FIG. 2 is an exemplary graph of historical lighting data;
FIG. 3 is a schematic view of the first division point for each day;
fig. 4 is a schematic view of the superposition of the first segmentation points for all days.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, a flowchart of a distributed intelligent building lighting adaptive energy-saving control method according to an embodiment of the present invention is shown, where the method includes the following steps:
and S001, acquiring historical illumination data, illumination sensor information and energy-saving lamp information in the building.
The present embodiment aims at performing adaptive energy-saving control on building illumination, and therefore, firstly, illumination conditions inside a building need to be analyzed, and in an actual situation, the illumination inside the building is mainly divided into brightness caused by ambient illumination and brightness caused by interior illumination, so that the influence of the ambient illumination must be considered in calculating control parameters of the lighting device, where the control parameters include which lighting devices are turned on and the brightness is turned on. Influenced by the building structure, some areas are in areas with bright ambient light for a long time in a day, and some areas are influenced by the reflection of light of windows and the like, sometimes in a darker condition and sometimes in a brighter condition; the sun is in the east-rising west-falling every day, and is influenced by the building position and the building design structure, and the environment illumination conditions of different areas of different buildings are different and have respective regularity. The internal illumination is usually constant, and the illumination change inside the building is mainly caused by the ambient illumination change, so the ambient illumination rule can be obtained by analyzing the illumination change rule inside the building. After the ambient light law of the building is obtained, the control parameters of the lighting devices in different areas in the building can be calculated according to the ambient light law, and the optimal energy-saving parameter adjustment control method is obtained.
For analyzing the illumination condition inside the building, specifically, installing illumination sensors in different areas in the building to obtain historical illumination data of the different areas, wherein each area is provided with a plurality of illumination sensors, the historical illumination data comprises illumination data of a plurality of days, the illumination data of each day consists of the illumination data collected by a plurality of time nodes, and the illumination data of each day is used as a period of the historical illumination data; meanwhile, the information of each illumination sensor comprises first position information of the illumination sensor, all the illumination sensors are arranged according to the first position information to form a sensor sequence, and the sensor sequence is obtained by arranging the illumination sensors in rows according to the floor from low to high and the same floor in sequence; the information of each lighting device, namely the energy-saving lamp, comprises the second position information of the energy-saving lamp and the relevant working parameters of the energy-saving lamp, such as the light intensity when the energy-saving lamp is used for lighting, and is recorded as the first working parameter. Referring to fig. 2, the illumination data collected by different sensors on the first day in the historical illumination data is shown, wherein a, B, C, D, E, F, G, H, \ 8230, the illumination sensors represent different areas in the building, each area has a plurality of illumination sensors, adjacent letter sensors are close in position, t1, t2, \ 8230, tn represent different time nodes; it should be noted that the illumination data of each day is used as the period of the historical illumination data, because the illumination data at the same time node of each day is similar, and each day conforms to the periodic variation rule of most ambient illumination.
It should be further noted that, the distributed intelligent lighting control system is that the system works according to a plurality of preset basic states, and the states can be mutually and automatically switched according to set time; correspondingly, in the embodiment, different working states corresponding to different time periods are calculated. The existing central centralized control system concentrates all control functions on a central processing unit, and once the central processing unit fails, the whole system is in a breakdown state, so that the system reliability is poor; the distributed control system distributes the control function to each module in the system, a central processing unit is not arranged, all the function modules can directly communicate with each other through a network bus, preset parameters are stored in each module, when a certain module in the system fails, only the function related to the module fails, and if a loop where the module is located fails, the normal work of other loops cannot be influenced; correspondingly, in the embodiment, the area with the similar ambient light change is calculated as the module.
Step S002, according to the historical illumination data of each sensor, a plurality of groups of regions with similar ambient illumination changes are obtained as a second category.
It should be noted that the illumination conditions of each day are not completely consistent, and the relationship between the illumination data of each day collected by each sensor is only similar but not completely equal, so that the DTW algorithm is used to calculate the similarity relationship between the illumination data of each sensor in each day to determine the segmentation points, and the sensors are divided into groups, so that the region corresponding to the sensor in each group is the region with similar ambient illumination change.
Specifically, the illumination data of a plurality of time nodes of each day collected by each sensor is arranged according to a time sequence to serve as a first illumination change sequence, and each sensor corresponds to one first illumination change sequence every day; it should be noted that, when the similarity between sequences is determined according to the conventional cosine similarity, the context of the elements is not limited, and in the illumination change sequence, the time sequence change is important and cannot be ignored, so the DTW algorithm is selected and used in the embodiment to analyze the relationship between the first illumination change sequences of the adjacent sensors.
Further, the DTW distance between the first illumination change sequences of every two adjacent sensors in the same day is obtained through calculation and is recorded as a first distance, the smaller the first distance is, the more similar the sequences are, after all the first distances in the same day are obtained, the frequency of occurrence of each first distance is calculated, the obtaining method is the ratio of the number of times of occurrence of a certain first distance to the number of all the first distances, and the ratio obtained here in this embodiment retains two decimal numbers; based on the first distance and the frequency of occurrence of the first distance,obtaining division threshold values of all first distances in the same day through OTSU threshold segmentation algorithm
/>
Selecting a value greater than a dividing threshold>
/>
Dividing the sensor sequence according to the adjacent sensor positions corresponding to the distances, wherein the divided positions are first dividing points, and each first dividing point corresponds to one first distance; in the plurality of first distances, each first distance is obtained by a first illumination change sequence of an adjacent sensor, each first distance corresponds to the position between two adjacent sensors, the plurality of first distances larger than a threshold value correspond to the positions between the adjacent sensors, the division of the sensor sequences is completed, and the sensor sequences are divided every day to obtain a plurality of initial categories; the first cut point was obtained for all days as described above.
It should be further noted that, if the superimposed segmentation points are close in the superimposed representation of the classification result of each day, the probability that the same change of different regions in the two classified classifications is caused by contingency is relatively small; otherwise, if the distance between the two superimposed classification division points is large, the large probability is caused by the contingency, the division points need to be removed, the division points existing in different days are reserved, and the reserved division points can represent the division of different classes.
Referring to fig. 3 and 4, fig. 3 shows the division of the first division points of each day, and fig. 4 is a representation of the first division points of all days after being superimposed; it is easy to find that, if a1, a2, a3 are close, there is a high possibility that a division point exists in the vicinity, b1, b2, b3 are close, there is a high possibility that a division point exists in the vicinity, and there is no close value around c1 and c3, so that these two values have a low possibility of being division points, and in fig. 4, b1, b2, b3 belong to the same position.
Specifically, the first division points of all the days are superimposed according to the position relationship of the sensor sequence, the superimposing method can refer to fig. 3 and fig. 4, that is, the division positions of the first division points of each day on the sensor sequence are superimposed into the same sensor sequence to obtain a division point sequence, each element in the division point sequence is a first division point, and the second distances of all the first division points in the division point sequence form a division point distance sequence; in this case, the closer the second distance of the first division point in the division point distance sequence is, the more likely these close first division points are to be division points for positions between the same adjacent sensors in the sensor sequence.
The method for acquiring the second distance of the certain first division point comprises the following steps: in a sensor sequence, the first division point is located between two sensors, the sensor sequence is arranged according to first position information of the sensors, first distances between all adjacent sensors on the days where the first division point is located are obtained, the sum of the first distances of all adjacent sensors on the left side of the position and the first distance of the first division point in the sensor sequence arranged from left to right is calculated before the position where the first division point is located, the sum is the sum of the first distances of the first division point, and the sum is recorded as a second distance of the first division point;
further, performing OTSU multi-threshold segmentation on the obtained segmentation point distance sequence to obtain a plurality of first categories, wherein each first category comprises second distances corresponding to a plurality of first segmentation points, the segmentation results of the plurality of first segmentation points in the same category as the segmentation points are similar, and the category reliability of each first category is used for representing the possibility that the corresponding position of the category as the segmentation point; to obtain to
/>
A first kindIn other examples, the class confidence level thereof>
/>
The calculation method comprises the following steps:
wherein, the first and the second end of the pipe are connected with each other,
/>
indicates the number of days, in this embodiment @, in the historical lighting data>
/>
Value is taken>
/>
，/>
/>
Indicates the fifth->
/>
Number of elements in a first category>
/>
Indicates the fifth->
/>
An element variance value in a first category, based on a variance value in a first category>
/>
Expressing an exponential function with a natural constant as a base; the elements in the first category are second distances of first segmentation points on different days, and normalization processing is performed at the same time in order to avoid the situation that the variance is 0, so that exponential function mapping with a natural constant as a base is adopted.
At this time, the process of the present invention,
/>
and the number of days>
/>
The greater the ratio of (A), the greater the value of>
/>
The closer to 1, the more days in all the days, the more the position in the division result is used as a division point for division, and the higher the credibility of the category as the division point is; variance value of elements
/>
The smaller the first distance corresponding to each first segmentation point in the category is, the moreThe closer, the higher the confidence of the class as a segmentation point.
Further, the category credibility of each first category is matched with a first preset threshold value
/>
Comparing, reserving a first category which is larger than a first preset threshold value, and then taking a first dividing point corresponding to a category center in a plurality of reserved first categories as a second dividing point representing the first category, wherein the second dividing point divides different sensors into a plurality of second categories according to the sensor sequences, and the dividing method is the same as that of the first dividing point for the sensor sequences; each second category comprises a plurality of illumination sensors, and illumination data collected by the illumination sensors have similar changes; the center of each first class is the center of the class which is the element with the largest sum of the ratios of the second distance to other second distances in the class, namely the second distance with the largest sum of the ratios of the second distance to other second distances in the class.
And S003, acquiring a second illumination change sequence of each sensor in the second category, and obtaining a third illumination change sequence of the second category through cosine similarity.
Dividing areas corresponding to different illumination sensors into different second categories, wherein illumination change conditions of illumination data in the same second category in one day are similar, and the same control mode should be adopted; the control mode refers to the change frequency of the lighting equipment, for the second type of illumination data with small illumination change, the change adjustment times of the lighting equipment in one day are less, and for the second type of illumination data with frequent illumination change, the change adjustment times of the lighting equipment in one day are more, so that different second type of illumination data are divided into different blocks for control; the same second type of illumination data corresponds to the same module, and different modules are not affected with each other, that is, the failure of one module does not affect another module.
It should be noted that, since each sensor in the same second category has a first illumination change sequence of multiple days, and there are multiple sensors in the second category, each first illumination change sequence of each sensor is similar, and the illumination change of each sensor is also similar, in order to reduce the amount of calculation, it is necessary to obtain an illumination change sequence that can represent the second category, which is called a third illumination change sequence.
Specifically, a second illumination variation sequence of a sensor in the same second category is obtained first, and is obtained by weighting the first illumination variation sequences of the sensor on all days, for example, any sensor in any second category is taken as an example, the second illumination variation sequence of the sensor is the first illumination variation sequence
/>
Point illumination data->
/>
The calculation method comprises the following steps:
wherein the content of the first and second substances,
/>
representing a number of days in historical lighting data>
/>
Indicates that the sensor is on >>
/>
The fifth ^ in the first sequence of illumination changes of the day>
/>
The illumination data of the point is taken>
/>
Indicates that the sensor is up to->
/>
A first weight value of a first sequence of lighting changes of a day; wherein->
/>
,/>
/>
Represents a fifth or fifth party>
/>
A second distance ratio of a first division point on the left side of the initial category divided by the sensor to a second division point on the left side corresponding to the first category to which the first division point belongs;
/>
represents a fifth or fifth party>
/>
A second distance ratio of a first dividing point on the right side of the initial category divided by the sensor to a second dividing point on the right side corresponding to the first category to which the first dividing point belongs; it should be noted that the ratio of the second distances between the first dividing point and the second dividing point is small to large, i.e. the ratio results are all smaller than 1.
At this time, the closer the two first division points on the left and right of the illumination area corresponding to each sensor on a certain day are to the second division points belonging to the second category, the closer the condition of the illumination area in the day is to the general condition, and when the value of each illumination area is calculated through the illumination data of multiple days, the greater the weight of the illumination data of the day should be, that is, each illumination data in the second illumination change sequence of each sensor is obtained through a weighted summation method.
Further, the cosine similarity between the second illumination change sequences of each sensor in each second category is calculated, the second illumination change sequence with the largest sum of the cosine similarities with other second illumination change sequences in each second category is used as the third illumination change sequence of each second category, the third illumination change sequence can represent illumination data change in a group of areas with similar illumination changes in one day, and the calculation amount required for obtaining the second control parameter is reduced.
And then, obtaining a plurality of reserved first categories according to the confidence degrees of the first categories, taking the center of each reserved first category as a second segmentation point, dividing the sensor sequences to obtain a plurality of second categories, weighting all the first illumination change sequences of each sensor in the second categories to obtain a second illumination change sequence of each sensor, and obtaining a third illumination change sequence of each second category according to the cosine similarity between the second illumination change sequences of each sensor.
And step S004, dividing each third illumination change sequence into a plurality of third categories through multi-threshold segmentation, and obtaining a second control parameter for each third category.
Dividing the third illumination change sequence of each second category into a plurality of parts by an OTSU multi-threshold segmentation method, wherein each part is marked as each third category, the illumination data in each third category are similar, and the illumination data difference between different third categories is large, so that the illumination data in the same third category are used as the illumination data of the same working mode, and the same control parameter can be adopted for illumination equipment adjustment in the same working mode and are marked as a second control parameter; the same working mode indicates that the illumination data of the corresponding areas of all the sensors in the second category to which the sensors belong are similar under the working mode.
Specifically, compensation light in the same working mode is calculated, namely, under the current ambient illumination, partial brightness needs to be increased through lighting equipment, so that the human visual sensation in an illumination area is good; constructing an environment light graph structure, wherein a node of the graph structure is each sensor in a second category, node data is illumination data of the same third category of each sensor, namely the illumination data in the same working mode, and the Euclidean distance between each sensor is used as an edge of the graph structure; furthermore, the illumination data which makes the human body have good visual sensation is recorded as
/>
Comparing each node data with ^ er>
Calculating a difference by subtracting/combining the node data, replacing the difference with each node data>
/>
If the result is less than 0, the node data is 0, and the obtained compensation optical diagram structure in the same working mode of a certain second category, namely a compensation optical diagram structure of a certain third category, is obtained at this moment.
Further, an energy-saving lamp graph structure is constructed according to second position information of all energy-saving lamps (lighting equipment) in the area corresponding to the sensor included in the second category, the node is each energy-saving lamp, and the Euclidean distance between the energy-saving lamps is defined as the edge; connecting a compensation light graph structure with an energy-saving lamp graph structure through first position information and second position information to form an abnormal graph, wherein the edge values of energy-saving lamp nodes and sensor nodes on the abnormal graph are still Euclidean distances, the obtained abnormal graph is used as the input of a neural network of the abnormal graph structure to be trained, the obtained output result is a vector, the value range of each element in the vector is [0,1], each element represents the percentage of the actual brightness of each lamp corresponding to the maximum brightness, and the percentage is the compensation light parameter of the brightness required to be provided by each energy-saving lamp; for example: [0.1,0.7,0.6] indicates that the brightness of the corresponding lamp is 0.1 times, 0.7 times, and 0.6 times the maximum brightness, respectively.
Further, according to the compensation light parameter of each energy-saving lamp and the first working parameter of the energy-saving lamp, a second control parameter of a third category is obtained, and the second control parameter obtaining method comprises the following steps: taking the product of the compensating light parameter and the first working parameter of each energy-saving lamp in the same third category as the first control parameter of each energy-saving lamp, wherein the first working parameter is the working power of the energy-saving lamp for starting the maximum brightness, the first control parameter is the working power of each energy-saving lamp in the same third category, and the first control parameter of each energy-saving lamp in the same category constitutes the second control parameter of the third category; and acquiring second control parameters of all third categories in each second category according to the method.
The heterogeneous graph structure neural network is an NARS network, and the method in the embodiment is used to calculate a large number of compensation light graph structures inside the building, and construct a corresponding heterogeneous graph structure by combining the energy-saving lamp graph structure inside each building, so as to obtain a heterogeneous graph structure data set. The corresponding labels are tested and scored by experts, and when the brightness of each energy-saving lamp in each building is considered to be the maximum brightness ratio, the energy-saving effect is the best while the requirement of saturated light is met.
And S005, controlling the energy-saving lamp in the building according to the second control parameter.
And when the area corresponding to a certain sensor is switched to the working mode of one of the second categories, adjusting and controlling the energy-saving lamp in the area according to the corresponding second control parameter, so as to realize the self-adaptive control of the lighting equipment (the energy-saving lamp) in the building by the distributed sub-modules.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (5)
1. A distributed intelligent building illumination self-adaptive energy-saving control method is characterized by comprising the following steps:
acquiring first position information of an illumination sensor inside a building and second position information of an energy-saving lamp;
acquiring a first illumination change sequence of each sensor every day, acquiring a first distance between the first illumination change sequences of adjacent sensors and the occurrence frequency of each first distance, acquiring first division points of all the first illumination change sequences every day according to the occurrence frequency of all the first distances, superposing the first division points of all the days, acquiring a second distance of each first division point, acquiring a plurality of first categories by multi-threshold division of the second distances, acquiring category credibility according to the second distance of the first division points in each first category, comparing the category credibility with a first preset threshold to acquire a plurality of second division points, and dividing all the sensors according to the second division points to acquire a plurality of second categories;
acquiring a first weight value of each first illumination change sequence according to the first division point and the second division point, weighting illumination data of the first illumination change sequence of each sensor according to the first weight value to obtain a second illumination change sequence of each sensor, and acquiring a third illumination change sequence of each second category according to cosine similarity between the second illumination change sequences of each sensor in each second category;
dividing each third illumination change sequence by multi-threshold segmentation to obtain a plurality of third categories, obtaining a compensation light map structure according to illumination data in the third categories and first position information, obtaining an energy-saving lamp map structure according to second position information, connecting the compensation light map structure and the energy-saving lamp map structure according to the first position information and the second position information to form an abnormal composition, and obtaining a second control parameter of the third categories according to the abnormal composition;
controlling the energy-saving lamp in the building according to the second control parameter;
the method for acquiring the category credibility comprises the following specific steps:
wherein the content of the first and second substances,
indicates the fifth->
A class confidence of a first class>
Indicates a number of days in the historical lighting data, and->
Is shown as
The number of elements in the first category indicates the ^ th ^ or ^ th>
An element variance value in a first category, based on a value of a square or a square>
Expressing an exponential function with a natural constant as a base; the elements in the first category are second distances of first segmentation points on different days;
the method for acquiring the compensation light pattern structure comprises the following specific steps:
taking the illumination sensor in any one second category and illumination data of different third categories in the second category as node information, taking the distance acquired by the first position information as an edge of a graph to form an environment light graph structure, and replacing the illumination data in the node information with a difference value between saturated light intensity to acquire a compensation light graph structure; the saturated light intensity represents a light intensity that makes human visual perception good;
the method for acquiring the energy-saving lamp diagram structure comprises the following steps: and constructing an energy-saving lamp graph structure according to second position information of all energy-saving lamps in the area corresponding to the sensor included in the second category, wherein the nodes are all the energy-saving lamps, and the edges are Euclidean distances among the energy-saving lamps.
2. The distributed intelligent building lighting adaptive energy-saving control method according to claim 1, wherein the first illumination change sequence of each sensor per day represents illumination data obtained by dividing historical illumination data collected from each sensor in a period of each day, each day illumination data of each sensor comprises illumination data collected by a plurality of time nodes, and the illumination data of the time nodes form an illumination data sequence, namely the first illumination change sequence of each sensor per day.
3. The method according to claim 1, wherein the obtaining of the first division points of all the first illumination variation sequences per day according to the occurrence frequency of all the first distances comprises:
according to the frequency of the first distances, obtaining a division threshold of the first distances through an OTSU threshold division algorithm to obtain a plurality of first distances larger than the division threshold, dividing the sensor sequence according to the first distances, wherein the corresponding division points are first division points of all first illumination change sequences every day, and the first division points divide the sensors every day into a plurality of initial categories;
the sensor sequence is to acquire the sensor according to the first position information.
4. The method according to claim 1, wherein the second illumination variation sequence of each sensor is obtained by:
wherein the content of the first and second substances,
indicates the ^ th or ^ th lighting change sequence of a certain sensor>
The illumination data of the point is taken>
Representing a number of days in historical lighting data>
Indicates that the sensor is up to->
Illumination data of a first point in a first sequence of illumination changes of day @>
A first weight value representing a first sequence of illumination changes for the sensor day k.
5. The method according to claim 4, wherein the specific method for obtaining the first weight value of the first illumination variation sequence is:
the first weight value is obtained from the similarity mean value of left and right first segmentation points of an initial category to which a certain first illumination change sequence belongs and left and right second segmentation points of a second category to which the first illumination change sequence belongs.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211588933.8A CN115696690B (en) | 2022-12-12 | 2022-12-12 | Distributed intelligent building illumination self-adaptive energy-saving control method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211588933.8A CN115696690B (en) | 2022-12-12 | 2022-12-12 | Distributed intelligent building illumination self-adaptive energy-saving control method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115696690A CN115696690A (en) | 2023-02-03 |
| CN115696690B true CN115696690B (en) | 2023-04-04 |
Family
ID=85055120
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211588933.8A Active CN115696690B (en) | 2022-12-12 | 2022-12-12 | Distributed intelligent building illumination self-adaptive energy-saving control method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115696690B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116257735B (en) * | 2023-05-16 | 2023-07-25 | 保定思齐智科信息科技有限公司 | Data processing method and system for smart city management |
| CN116705180B (en) * | 2023-08-08 | 2023-10-31 | 山东北国发展集团有限公司 | N2O catalytic decomposition monitoring method and system based on multidimensional data analysis |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102196615A (en) * | 2010-03-01 | 2011-09-21 | 柳延东 | Environmentally-adaptive energy-saving controller of lighting devices in classroom |
| CN104238510A (en) * | 2014-09-03 | 2014-12-24 | 尚高科技有限公司 | Intelligent building district management system based on Internet of things |
| CN106028525A (en) * | 2016-06-21 | 2016-10-12 | 泰华智慧产业集团股份有限公司 | Differential evolution algorithm based adaptive indoor dimming method and system |
| WO2021072951A1 (en) * | 2019-10-14 | 2021-04-22 | 苏州数言信息技术有限公司 | Internet of things and machine learning-based smart indoor energy conservation system |
| CN114205973A (en) * | 2021-11-12 | 2022-03-18 | 杭州科技职业技术学院 | Intelligent control system for public place illumination |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012061709A1 (en) * | 2010-11-04 | 2012-05-10 | Digital Lumens Incorporated | Method, apparatus, and system for occupancy sensing |
| CN102645926A (en) * | 2012-05-11 | 2012-08-22 | 山东理工大学 | Intelligent energy-saving control system for classroom lighting and fans |
| CN104066220B (en) * | 2013-03-19 | 2016-06-08 | 日电（中国）有限公司 | A kind of illumination control method and device |
| CN110582149A (en) * | 2019-08-21 | 2019-12-17 | 安徽正风建设工程有限公司 | Energy-saving automatic conference light regulation and control system |
| CN110673475B (en) * | 2019-09-17 | 2020-12-18 | 重庆特斯联智慧科技股份有限公司 | Intelligent building energy-saving system with internet of things sensing function and control method |
| CN111148323B (en) * | 2019-12-25 | 2022-07-19 | 重庆特斯联智慧科技股份有限公司 | Intelligent building illuminating lamp light control method and system based on target perception |
| CN114980450B (en) * | 2022-08-02 | 2022-10-25 | 利亚德光电股份有限公司 | Regional light environment online evaluation display method, system, medium and equipment |
-
2022
- 2022-12-12 CN CN202211588933.8A patent/CN115696690B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102196615A (en) * | 2010-03-01 | 2011-09-21 | 柳延东 | Environmentally-adaptive energy-saving controller of lighting devices in classroom |
| CN104238510A (en) * | 2014-09-03 | 2014-12-24 | 尚高科技有限公司 | Intelligent building district management system based on Internet of things |
| CN106028525A (en) * | 2016-06-21 | 2016-10-12 | 泰华智慧产业集团股份有限公司 | Differential evolution algorithm based adaptive indoor dimming method and system |
| WO2021072951A1 (en) * | 2019-10-14 | 2021-04-22 | 苏州数言信息技术有限公司 | Internet of things and machine learning-based smart indoor energy conservation system |
| CN114205973A (en) * | 2021-11-12 | 2022-03-18 | 杭州科技职业技术学院 | Intelligent control system for public place illumination |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115696690A (en) | 2023-02-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115696690B (en) | Distributed intelligent building illumination self-adaptive energy-saving control method | |
| CN111864896B (en) | Power load monitoring method and system | |
| CN110673475B (en) | Intelligent building energy-saving system with internet of things sensing function and control method | |
| CN109005628A (en) | Street lamp color temperature adjusting method, street lamp and computer readable storage medium | |
| CN113325762B (en) | Intelligent building personalized energy consumption control method, system, device and equipment | |
| CN106499656B (en) | A kind of fan wind speed intelligent control method | |
| CN111191687B (en) | Power communication data clustering method based on improved K-means algorithm | |
| CN109389212A (en) | A kind of restructural activation quantization pond system towards low-bit width convolutional neural networks | |
| CN108171241B (en) | IFCS/Otsu-based IOFR flame identification method | |
| CN109799726B (en) | Smart home system combined with living environment detection | |
| CN116321610B (en) | Intelligent light control method and control system | |
| CN107944617A (en) | A kind of doubtful stealing theme influence factor weight optimization method that logic-based returns | |
| CN115968088A (en) | Intelligent tunnel dimming method and system and computer storage medium | |
| CN113923839A (en) | Light adjusting method and system | |
| CN108282943A (en) | Illumination control method and device | |
| CN114727461A (en) | Light control system based on intelligent home | |
| CN116561535B (en) | Individualized building interaction design processing method | |
| CN114091758A (en) | Medium-short term load prediction method, system and storage medium | |
| CN116740809B (en) | Intelligent sofa control method based on user gesture | |
| CN116470491A (en) | Photovoltaic power probability prediction method and system based on copula function | |
| CN113726559B (en) | Based on artificial intelligence network safety analysis early warning system | |
| CN110409987B (en) | Photosensitive shutter self-adaptive control method and system | |
| Li et al. | Image object detection algorithm based on improved Gaussian mixture model | |
| CN117762032B (en) | Intelligent equipment control system and method based on scene adaptation and artificial intelligence | |
| CN112541397A (en) | Flame detection method based on improved ViBe algorithm and lightweight convolutional network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |