Indoor illumination regulation and control method and system based on illumination gradient

Abstract

The invention belongs to the technical field of illumination control, and provides an indoor illumination regulation and control method and system based on illumination gradient, and the method comprises the steps: collecting indoor dynamic light environment data; calculating an indoor illumination gradient according to the light environment data; predicting indoor light satisfaction on the basis of the obtained illumination gradient; and performing combination optimization of the curtain opening and the lamp gear according to the light satisfaction prediction value, and dynamically adjusting the curtain and the lamp in combination with a combination optimization result to complete indoor illumination regulation and control based on the illumination gradient.

Description

Technical Field

[0001] The present invention belongs to the technical field of lighting control, and in particular relates to an indoor lighting control method and system based on illumination gradient. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] The quality of the indoor light environment affects users' visual comfort and usage efficiency. Currently, indoor lighting mainly relies on the simple superposition of natural light and artificial light. However, there are the following problems in indoor lighting control: (1) Uneven illumination distribution. That is, when the interior space is deep, natural light attenuates significantly from the exterior windows to the interior windows, resulting in large differences in desktop illumination in different areas. (2) Lack of personalized needs, that is, the lighting environment cannot be dynamically adjusted according to individual preferences, and the difference in light satisfaction is significant; (3) Energy waste, that is, the lack of coordinated control between artificial lighting and natural light, often leading to over-lighting or under-lighting; (4) The adjustment means are single, that is, the adjustment of curtains and lamps relies on manual operation and lacks intelligent dynamic optimization strategies.

[0004] Currently, indoor lighting control typically relies on fixed sensors to adjust the illumination of lamps. However, this approach is unable to adapt to complex changes in indoor light environments and does not address personalized needs. Therefore, there is an urgent need for intelligent control systems that can sense, predict, and optimize indoor light environments in real time to achieve efficient synergy between natural and artificial light. Summary of the Invention

[0005] To solve the above problems, the present invention proposes an indoor lighting control method and system based on illumination gradient, which dynamically collects light environment data, calculates the indoor illumination gradient, predicts light satisfaction based on the illumination gradient, optimizes the indoor curtain opening and lamp gear according to the light satisfaction prediction value, completes indoor lighting control based on illumination gradient, and realizes personalized and energy-efficient indoor lighting.

[0006] According to some embodiments, a first solution of the present invention provides an indoor lighting control method based on illumination gradient, which adopts the following technical solutions: A method for controlling indoor lighting based on illumination gradient, comprising: Collect dynamic light environment data indoors; Calculate the indoor illumination gradient based on the light environment data; Based on the obtained illuminance gradient, predict the indoor light satisfaction; The curtain opening and lamp level are optimized according to the light satisfaction prediction value. The curtains and lamps are dynamically adjusted based on the combined optimization results to complete the indoor lighting control based on the illumination gradient.

[0007] As a further technical limitation, before calculating the indoor light satisfaction, the collected dynamic light environment data is processed by interpolation method to obtain the illuminance distribution map at different indoor locations; in the obtained illuminance distribution map, the illuminance value decreases with increasing distance from indoor and outdoor windows.

[0008] As a further technical definition, the illumination gradient is the ratio of the difference between the illumination of the center point of the indoor desktop and the illumination of each edge point to the distance between the center point and the edge point; on the horizontal cross-section of the room, the illumination gradients in different directions are defined with the intersection of the cross-section diagonals as the center and the diagonals and median lines passing through the center as the directions.

[0009] Furthermore, the illumination gradients in different directions are calculated to detect outliers, and the detected outliers are eliminated by combining the interquartile range method. The illumination gradients in different directions after eliminating the outliers are analyzed, and the increase or decrease of illumination in that direction is determined based on the positive or negative value of the illumination gradient.

[0010] As a further technical limitation, a multi-layer perceptron model is used to predict light satisfaction. The multi-layer perceptron used includes an input layer, an output layer, and at least one hidden layer. It learns the nonlinear characteristics of the illuminance gradient in a feedforward manner and completes the prediction of light satisfaction based on the classification and regression of the illuminance gradient.

[0011] As a further technical limitation, in the process of optimizing the combination of curtain opening and lamp level, a multi-objective optimization algorithm is used to optimize the combination of curtain opening and lamp level. The curtain opening is roughly adjusted according to the light satisfaction prediction value until the indoor light satisfaction value is higher than the satisfaction threshold. The lamp level is adjusted in combination with the indoor satisfaction value, and the indoor artificial illumination value is changed to obtain the optimal combination of curtain opening and lamp level. The curtain opening and lamp level are dynamically adjusted in real time according to the obtained optimal combination.

[0012] It should be noted that in the process of optimizing the combination of curtain opening and lamp level, a multi-objective optimization algorithm is used to optimize the combination of curtain opening and lamp level. The curtain opening is roughly adjusted according to the light satisfaction prediction value until the indoor light satisfaction level reaches more than 50% of "satisfactory". For unsatisfactory situations, lamp control is carried out, and the lamp level is adjusted in combination with the indoor satisfaction value to change the indoor artificial illumination value to obtain the optimal combination of curtain opening and lamp level. The curtain opening and lamp level are dynamically adjusted in real time according to the obtained optimal combination.

[0013] According to some embodiments, a second solution of the present invention provides an indoor lighting control system based on illumination gradient, which adopts the following technical solution: An indoor lighting control system based on illumination gradient, comprising: A collection module configured to collect dynamic light environment data in a room; A calculation module configured to calculate an indoor illumination gradient based on light environment data; a prediction module configured to predict indoor light satisfaction based on the obtained illumination gradient; The control module is configured to optimize the combination of curtain opening and lamp level according to the light satisfaction prediction value, and dynamically adjust the curtains and lamps based on the combined optimization results to complete the indoor lighting control based on the illumination gradient.

[0014] According to some embodiments, a third solution of the present invention provides a computer-readable storage medium, which adopts the following technical solution: A computer-readable storage medium stores a program, which, when executed by a processor, implements the steps of the indoor lighting control method based on illumination gradient as described in the first solution of the present invention.

[0015] According to some embodiments, a fourth solution of the present invention provides an electronic device, which adopts the following technical solution: An electronic device includes a memory, a processor, and a program stored in the memory and running on the processor. When the processor executes the program, the steps of the indoor lighting control method based on illumination gradient as described in the first embodiment of the present invention are implemented.

[0016] According to some embodiments, a fifth solution of the present invention provides a computer program product, which adopts the following technical solution: A computer program product includes software code, wherein the program in the software code executes the steps of the indoor lighting control method based on illumination gradient as described in the first embodiment of the present invention.

[0017] Compared with the prior art, the present invention has the following beneficial effects: The present invention dynamically collects light environment data, calculates indoor illumination gradients, predicts light satisfaction based on illumination gradients, and optimizes indoor curtain openings and lamp positions according to the predicted light satisfaction values, thereby achieving personalized and energy-efficient indoor lighting. This effectively solves the problem of lack of coordinated control of artificial lighting and natural light, and dynamically adjusts curtain openings and lamp positions based on light satisfaction, thereby improving the intelligence level of indoor lighting regulation. At the same time, through personalized adjustment combined with student satisfaction scores, "one person, one policy" light environment optimization is achieved, visual comfort is improved, natural light changes are responded to in real time, and lighting strategies are dynamically adjusted through prediction and optimization algorithms. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The drawings constituting a part of the specification of this embodiment are used to provide a further understanding of this embodiment. The schematic embodiments and descriptions of this embodiment are used to explain this embodiment and do not constitute an improper limitation on this embodiment.

[0019] Figure 1 This is a flow chart of the indoor lighting control method based on illumination gradient in Example 1 of the present invention; FIG2 (a) is a schematic diagram of the external structure of a classroom in the first embodiment of the present invention; FIG2( b ) is a schematic diagram of the internal structure of a classroom in the first embodiment of the present invention; FIG3 (a) is a schematic diagram of a grid fluorescent lamp in Embodiment 1 of the present invention; FIG3 ( b ) is a schematic diagram of a double-tube fluorescent lamp in Embodiment 1 of the present invention; Figure 4 Schematic diagram of the distribution of desktop measurement points in Example 1 of the present invention; FIG5 (a) is a desktop illumination distribution diagram of classroom R1 in Example 1 of the present invention; FIG5( b ) is a tabletop illumination distribution diagram of classroom R2 in Example 1 of the present invention; FIG5( c ) is a tabletop illumination distribution diagram of classroom R3 in Example 1 of the present invention; FIG5( d ) is a tabletop illumination distribution diagram of classroom R4 in Example 1 of the present invention; FIG5( e ) is a tabletop illumination distribution diagram of classroom R5 in Example 1 of the present invention; FIG5( f ) is a tabletop illumination distribution diagram of classroom R6 in Example 1 of the present invention; FIG5( g ) is a desktop illumination distribution diagram of classroom R7 in Example 1 of the present invention; FIG5(h) is a tabletop illumination distribution diagram of the R8 classroom in the first embodiment of the present invention; FIG5(i) is a tabletop illumination distribution diagram of the R9 classroom in the first embodiment of the present invention; FIG5( j ) is a desktop illumination distribution diagram of classroom R10 in Example 1 of the present invention; FIG5( k ) is a tabletop illumination distribution diagram of classroom R11 in Example 1 of the present invention; FIG5 (l) is a tabletop illumination distribution diagram of the R12 classroom in the first embodiment of the present invention; FIG5( m ) is a tabletop illumination distribution diagram of classroom R13 in Example 1 of the present invention; FIG5(n) is a tabletop illumination distribution diagram of classroom R14 in Example 1 of the present invention; Figure 6 This is a schematic diagram of lighting satisfaction results in Example 1 of the present invention; Figure 7Schematic diagram of illumination gradient distribution in embodiment 1 of the present invention; FIG8 (a) is a schematic diagram of the illumination gradient distribution of D1, D2 and D8 in the first embodiment of the present invention; FIG8( b ) is a schematic diagram of the illumination gradient distribution of D4 , D5 , and D6 in the first embodiment of the present invention; FIG8 (c) is a schematic diagram of the illumination gradient distribution of D3 and D7 in the first embodiment of the present invention; Figure 9 This is a flow chart of dynamic adjustment of curtains and lamps in the first embodiment of the present invention; Figure 10 This is a structural block diagram of an indoor lighting control system based on illumination gradient in the second embodiment of the present invention. DETAILED DESCRIPTION

[0020] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0021] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0022] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0023] In the present invention, terms such as "upper", "lower", "left", "right", "front", "back", "vertical", "horizontal", "side", "bottom", etc. indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings. They are relational words determined only for the convenience of describing the structural relationships of the various parts or elements of the present invention, and do not specifically refer to any part or element in the present invention, and should not be understood as limiting the present invention.

[0024] In the present invention, terms such as "fixed connection," "connected," and "connection" should be interpreted broadly to mean a fixed connection, an integral connection, or a detachable connection; a direct connection or an indirect connection through an intermediary. Relevant researchers or technicians in this field may determine the specific meanings of these terms in the present invention based on specific circumstances, and they should not be construed as limitations of the present invention.

[0025] In the absence of conflict, the embodiments of the present invention and the features thereof may be combined with each other.

[0026] Example 1 The first embodiment of the present invention introduces an indoor lighting control method based on illumination gradient.

[0027] like Figure 1 The indoor lighting control method based on illumination gradient shown in FIG. 1 includes: Collect dynamic light environment data indoors; Calculate the indoor illumination gradient based on the light environment data; Based on the obtained illuminance gradient, predict the indoor light satisfaction; The curtain opening and lamp level are optimized according to the light satisfaction prediction value. The curtains and lamps are dynamically adjusted based on the combined optimization results to complete the indoor lighting control based on the illumination gradient.

[0028] Below, this embodiment takes a primary school classroom as an example to provide a detailed introduction to indoor lighting control.

[0029] Taking into account the impact of floors on the indoor light environment, this embodiment selects 14 classrooms (defined as R1, R2, R3, ..., R14) with completely consistent decoration layouts on different floors. The exterior structure schematic diagram and the interior structure schematic diagram of the selected classrooms are shown in Figure 2 (a) and Figure 2 (b), respectively. The basic parameters of the classrooms are shown in Table 1, among which the grid fluorescent lamp is shown in Figure 3 (a), the double-tube fluorescent lamp is shown in Figure 3 (b), the curtain status is partially used, indicating that half of the windows have curtains, the desk height is set to 700 mm, and the power of the lighting in the classroom is set to 9.6 watts / square meter.

[0030] Table 1 Basic parameters of classroom

[0031] The light environment of the classroom, including illuminance and distribution, was measured using a lux meter on the morning of the experimental day when students were studying in the classroom to simulate real conditions. The indoor lighting mode of the classroom was a combination of natural light and artificial lighting, and all lamps were kept on during the measurement.

[0032] The desktop is the main area for students’ visual work; according to the daylight measurement method and taking into account the actual layout of desks and chairs in the classroom, this embodiment measures the measurement points as follows: Figure 4 The layout shown is as follows; each classroom has 12 measuring points, named from A to L; the measuring points are horizontally distributed in the center of the first, third and fifth rows of desks, with a height of 700 mm; the distance between the measuring points is 2000 mm front to back and 1900 mm left to right.

[0033] This embodiment establishes an illuminance mapping table corresponding to different curtain openings under natural light conditions, and associates the illuminance values ​​inside and outside the exterior windows. Specifically, under normal sunlight exposure, illuminance measuring instruments are installed at the following locations in the classroom: outside the exterior window, and at evenly spaced locations from the inside of the exterior window to the interior wall of the classroom (assuming that the distance from the exterior window to the interior wall of the classroom is 7m, a total of 8 measuring points are arranged at even intervals of 1m from the inside of the exterior window to the interior wall). The curtains arranged on the exterior windows are controlled to adjust the opening of the curtains between 0-100% and 10% in intervals of 10%. The illuminance values ​​at each measuring point are measured at different curtain openings, and a mapping relationship is established to facilitate subsequent adjustment of the curtain opening based on the prediction results.

[0034] In this embodiment, the classroom light environment satisfaction is evaluated based on the monitoring data of the light environment. The average illuminance on the desk surface should not be less than 300 lx. In order to fully reflect the illuminance distribution of the classroom desktop, the interpolation method is used to draw the illuminance distribution maps of different classrooms as shown in Figures 5(a), 5(b), 5(c), 5(d), 5(e), 5(f), 5(g), 5(h), 5(i), 5(j), 5(k), 5(l), 5(m) and 5(n), with the origin located in the center of the classroom.

[0035] Comparing the illuminance distribution across classrooms reveals that illuminance values ​​decrease with increasing distance from the exterior window. The first row of curtain-free classrooms near the window has higher illuminance, typically exceeding 2000 lx. The fourth row of desks near the inner corridor has the lowest average illuminance, accounting for 70% of those that fail to meet the standard. Excessive illuminance variations can easily cause visual fatigue in elementary school students, affect the attention of students near windows, and lead to low learning efficiency. This is primarily due to uneven outdoor lighting caused by varying distances between desks and exterior windows. This can be improved by adjusting the curtains. For example, in classroom R8, shown in Figure 5(h), using the curtains at the front of the classroom while keeping those at the back closed significantly improves the uniformity of the indoor lighting.

[0036] Lighting satisfaction results are as follows Figure 6 As shown in the figure, across 14 classes, an average of 90.2% of students were satisfied or even very satisfied with the lighting environment, while only 2.2% were dissatisfied. Classrooms with a higher proportion of dissatisfaction had lower average indoor illuminance, most below 500 lx. Analysis of the locations of students dissatisfied with the lighting environment revealed that they were primarily concentrated near exterior windows and corridors. Near windows, dissatisfied students believed the classroom lighting was too bright because the sunlight received was direct. Near corridors, dissatisfied students believed the classroom lighting was too dim because the sunlight received, which attenuates with increasing distance, was less intense.

[0037] Therefore, uneven lighting is a prominent feature of classrooms. To quantify the distribution of light across the entire field of view and the impact of dynamic changes in the light environment, this embodiment introduces an illumination gradient: the difference between the illumination at the center of the desk and at each edge point divided by the distance between the two points. As shown in Figure 7, starting from due south and rotating clockwise, eight illumination gradients are defined in sequence.

[0038] Before analyzing the illumination gradients obtained in different directions, in order to avoid the influence of factors such as experimental errors, outliers were first detected by calculating the Z value, and then the interquartile range method was used to eliminate outliers to further ensure the rationality of the data.

[0039] Figures 8(a)-8(c) show schematic diagrams of illuminance gradient distribution in different directions. Analysis of these illuminance gradients reveals that, as shown in Figure 8(a), at D1, D2, and D8, most illuminance gradients are negative. These three illuminance gradients indicate increasing illuminance, with average values ​​ranging from -30 to -60 lx / m. As shown in Figure 8(b), at D4, D5, and D6, most are positive, indicating decreasing illuminance, with average values ​​ranging from 30 to 50 lx / m. As shown in Figure 8(c), at D3 and D7, the distribution of positive and negative values ​​is relatively uniform, with slight illuminance variations, with average values ​​of -15 lx / m and 17 lx / m, respectively.

[0040] Spearman correlation analysis was used to examine the relationship between illuminance gradients and student light satisfaction. As shown in Table 2, illuminance gradient D3 showed a significant correlation with student light satisfaction. D3 points from the center of the desk toward the blackboard, falling within the student's direct field of vision when seated. Students can quickly perceive and react to changes in the light environment, which in turn affects their light satisfaction. However, while D1, D2, and D8 represent increasing illuminance values ​​and possess high illuminance values, their relatively small proportions within the student's primary field of vision do not significantly correlate with student light satisfaction. Similarly, while D4, D5, and D6 represent decreasing illuminance values ​​and possess high values, their limited impact within the primary field of vision does not result in a significant correlation. Furthermore, D7 points directly behind the student when seated and is also outside the student's field of vision, and therefore has no significant impact on light satisfaction.

[0041] Table 2 Correlation analysis results between different variables and light

[0042] This example uses field measurements of 14 classrooms to determine lighting parameters and student lighting satisfaction. This example uses a data-driven approach to predict student satisfaction with classroom lighting in real time. This algorithm is trained using a labeled dataset, enabling the trained algorithm to classify the data and accurately predict outcomes.

[0043] Input features were selected based on their correlation with the prediction target (i.e., student satisfaction with classroom lighting). Input features were primarily divided into two categories: First, considering that the illuminance at the center of the student's desk directly reflects the overall illuminance in the field of view, and that illuminance values ​​have been shown in previous studies to affect students' physical and psychological well-being, this was selected as the first category of input features. Second, the Spearman correlation coefficient between illuminance gradients in different directions and student light satisfaction was calculated. The results showed a significant correlation between illuminance gradients in different directions and student light satisfaction, so this was selected as the second category of input features. The dataset contains 156 data points, and these input features were used to predict the output feature, namely, student light satisfaction.

[0044] To avoid assigning different weights to values ​​of varying magnitudes during data analysis, all feature values ​​were normalized to the range [0, 1] before learning. In Dataset 1, there are two independent variables: the illuminance measurement at the center of the student's current location (Lm) and the illuminance gradient (D3), a key correlated variable selected through correlation analysis. There is also a dependent variable: the student's satisfaction with lighting, measured in real time. The dataset was split into training and test sets with an 85% to 15% split.

[0045] During the illuminance gradient prediction process, a relational table (file) is created for each desk in the classroom. This table contains three parameters: the center illuminance Lm of the desk (which can be directly measured), the illuminance gradient value D3 (calculated and stored according to a formula), and the student's light satisfaction level (the results will be stored for subsequent predictions). Initially, the relevant machine learning model is called to read the first two parameters as input features for prediction (the code can be used to read parameters at specific locations in the corresponding file for model prediction). The predicted results are then stored in the third parameter. After the prediction is complete, the light satisfaction prediction results for each desk are compared with the expected results, and the lighting environment is subsequently optimized and controlled based on the different comparison results.

[0046] It's important to note that the light satisfaction prediction is the result, and the illuminance gradient is the cause, similar to obtaining Y from X, and making subsequent changes based on the Y result. X may change due to the first Y, which in turn affects the second Y. Specifically, the first prediction calculates light satisfaction based on the directly measured Lm and calculated D3. If the lighting environment is subsequently adjusted, the corresponding Lm and D3 will also change based on the specific first light satisfaction prediction. A further prediction is then made based on the changed Lm and D3. Optimization and changes are continuously made until the light satisfaction result reaches the desired level.

[0047] To investigate whether illuminance gradients affect student satisfaction with classroom lighting, this example constructed Dataset 2 based on the predictions from the previous classification machine learning model. This dataset includes one variable: illuminance (Lm) measured at the center of the desks, and student satisfaction with lighting. While maintaining the same model parameters, the prediction accuracy of the two datasets was compared to explore whether illuminance gradients significantly affect lighting satisfaction.

[0048] Six different classification machine learning models were used for this dataset: gradient boosting trees, decision trees, K-nearest neighbors, random forests, AdaBoost, and multilayer perceptrons. For each model, the training-test ratio was 8.5:1.5, and the operational procedures remained the same. Gradient boosting trees (GBTs) exhibit strong comprehensive learning capabilities and excel when handling complex data. Decision trees (DTs) are easy to understand and visualize, making them suitable for preliminary exploration of data features. K-nearest neighbors (KNNs) are simple and intuitive, offering good interpretability and are suitable for small datasets. Random forests (RFs) improve model stability and overfitting resistance by constructing multiple decision trees. AdaBoost significantly improves model accuracy by combining multiple weak classifiers in a weighted manner. Multilayer perceptrons (MLPs) excel at handling nonlinear data and capturing complex patterns. By comparing and evaluating these models, the goal was to identify the optimal classification method for improving prediction accuracy.

[0049] 1) Gradient Boosted Tree (GBT): The core idea of ​​the gradient boosted tree is to use the negative gradient of the loss function in the current model as an approximation of the residual. Essentially, it fits a regression tree by performing a first-order Taylor expansion on the loss function. When the gradient boosted tree is used for a classification model, it is a gradient boosted decision tree (GBDT), an iterative decision tree algorithm also known as MART (Multiple Regression Tree). It constructs a set of weak learners (trees) and accumulates the results of multiple decision trees as the final prediction output. This algorithm effectively combines the concepts of decision trees and ensembles. In this example, a set of 50 weak learners is constructed, and the seed number of the random number generator is set to 5.

[0050] 2) Decision Tree: A decision tree is a model that uses a tree-like data structure to display decision rules and classification results. As an inductive learning algorithm, it uses various techniques to transform seemingly chaotic known data into a tree-like model that can predict unknown data. Each path from the root node (the attribute that contributes most to the final classification result) to a leaf node (the final classification result) represents a decision rule. In this example, the maximum depth of each tree is set to 3, and the number of seeds in the random number generator is set to 15.

[0051] 3) K Nearest Neighbor (KNN): The KNN algorithm is a very special machine learning algorithm because it does not involve a typical learning process. Instead, it uses training data to partition the feature vector space, and the resulting partitioning serves as the final algorithm model. A set of sample data, also known as the training set, is used. Each data point in the sample set has a label, meaning that the correspondence between each data point and its corresponding category is known. After inputting unlabeled data, each feature of this unlabeled data point is compared with the corresponding features of the data in the sample set. The classification label of the data point with the closest feature (the nearest neighbor) in the sample is then extracted. Generally, the first K most similar data points in the sample set are selected; this serves as the origin of K in the KNN algorithm. Typically, K is an integer no greater than 20. Finally, the category that appears most frequently among the K most similar data points is selected as the classification for the new data point. In this example, a K value of 7 is assumed.

[0052] 4) Random Forest: Building on the bagging ensemble constructed using decision trees as the base learner, random forests further introduce random attribute selection (i.e., random feature selection) during the decision tree training process. When selecting a partitioning attribute, a traditional decision tree selects the best attribute from the current node's attribute set (assuming there are d attributes). In contrast, for each node in the base decision tree, a subset of k attributes is randomly selected from that node's attribute set, and the best attribute from this subset is then selected for partitioning. The parameter k here controls the degree of randomness introduced: if k = d, the base decision tree is constructed identically to a traditional decision tree; if k = 1, a random attribute is selected for partitioning. In this example, 50 decision trees are constructed, and the number of seeds in the random number generator is set to 20.

[0053] 5) AdaBoost: AdaBoost stands for Adaptive Boosting. Its adaptability lies in the fact that samples misclassified by the previous base classifier are weighted higher, while samples correctly classified are weighted lower and used again to train the next base classifier. Furthermore, a new weak classifier is added in each iteration until a predetermined sufficiently low error rate or a predetermined maximum number of iterations is reached, at which point a final strong classifier is determined. This example uses 100 weak learners to train the AdaBoost model, and the random generator seed number is set to 15.

[0054] 6) Multilayer Perceptron (MLP): An important neural network model composed of multiple layers of perceptrons, also known as a deep neural network (DNN). By simulating the workings of neural networks in the human brain, MLPs enable complex information processing and pattern recognition. In the fields of machine learning and artificial intelligence, MLPs have become the foundation of many important algorithms. Perceptrons, the fundamental component of MLPs, consist of an input layer, one or more hidden layers, and an output layer. Each perceptron node is connected to nodes in the previous and next layers. By adjusting the weights and thresholds between nodes, the perceptron learns patterns in the input data. As input data passes through the perceptron, it sequentially passes through each layer, and after processing at each layer, the final output is obtained. The most significant feature of a multilayer perceptron is its multiple layers of neurons, which enables it to handle more complex problems. Compared to single-layer perceptrons, MLPs have stronger representation and generalization capabilities. Furthermore, by properly designing the number and size of hidden layers, MLPs can approximate any continuous function, enabling the solution of highly nonlinear problems. In this example, two hidden layers are specified for the model, with 50 and 25 neurons in each, respectively. The maximum number of iterations for model training is defined as 300, and the seed number of the random number generator is set to 5.

[0055] In the model prediction performance results, this example uses the `accuracy_score` function to calculate the classification accuracy. This function compares the true labels of the test set, y_test, with the labels predicted by the model, y_pred, and calculates the proportion of samples where the two labels match. The accuracy score is expressed in decimal form, representing the percentage of samples that are correctly predicted. The accuracy score is a metric used to evaluate classification model performance, measuring the proportion of correctly predicted samples.

[0056] The prediction performance results of the six classification machine learning models in predicting lighting satisfaction are shown in Table 3. The first column lists the six selected classification machine learning models, and the second and third columns show the accuracy results of the predictions on two different datasets.

[0057] Table 3 Prediction performance results

[0058] Table 3 shows that, in the comparison of prediction accuracy across different dataset types, dataset 1 consistently outperformed dataset 2 across all six classification machine learning models, with most models achieving significant improvements in prediction accuracy. This suggests that considering illuminance gradients in different directions around students can better predict their lighting satisfaction, indicating that these gradients do have an impact on student lighting satisfaction. Since illuminance represents the intensity of light and the degree to which a surface is illuminated, variations in illuminance can lead to noticeable contrasts between bright and dark areas. This contrast can cause eye fatigue and visual discomfort, especially with prolonged exposure to uneven lighting. Symptoms such as dry eyes, itchiness, and blurred vision can be exacerbated. Furthermore, uneven lighting can distract students, reducing their concentration and productivity. When the eyes must adapt to varying brightness levels, the brain shifts attention to process these visual stimuli, leading to inattention and negatively impacting student lighting satisfaction.

[0059] Table 3 also shows that, among all six classification machine learning models, the multilayer perceptron model achieved the highest accuracy in predicting lighting satisfaction in both Datasets 1 and 2. This demonstrates the numerous advantages of multilayer perceptrons in machine learning prediction. First, multilayer perceptrons are able to model complex nonlinear relationships, adapting to a variety of data distributions. Second, their multilayer structure automatically learns and extracts important features from the data, reducing the need for manual feature engineering. Furthermore, multilayer perceptrons have strong expressiveness, can handle high-dimensional data, and exhibit good generalization capabilities. The adaptability of multilayer perceptrons makes them suitable for a variety of tasks, including classification and regression. Furthermore, by using the backpropagation algorithm, they can effectively optimize network weights, improving training efficiency. Finally, their scalability enables them to handle large datasets, making them ideally suited for big data applications.

[0060] Through a multi-objective optimization algorithm, the curtain opening and lamp gear combination is optimized to find the combination that minimizes energy consumption and maximizes light satisfaction. The specific process is as follows: prediction is made through the set model parameters and selected physical quantities. For the prediction results, if the proportion of unsatisfactory results is high, first determine whether the distribution of natural light in the classroom has caused a serious uneven illumination distribution. When large-scale adjustments to the illumination value are required, priority is given to changing the illumination value by controlling the opening and closing degree of the external window curtains. Read the illumination mapping table corresponding to the aforementioned different curtain openings, and determine to what opening the curtains should be controlled that the indoor illumination value is distributed within a relatively reasonable range, so that the illumination distribution is relatively uniform (coarse adjustment, for the extremely unreasonable illumination distribution of the light environment inside the classroom, a large range of indoor illumination values ​​needs to be changed). After adjusting the opening, collect illumination data from multiple measuring points on the desktop in real time and make a prediction, and judge the satisfaction ratio of the results. If adjustment is still needed (fine adjustment, for individuals with unsatisfactory prediction results, individual adjustment is made, then by changing the different gears of the desk lamp on the desk with poor prediction results, change the artificial illumination value, and combine the energy consumption required by the desk lamp at different gears. Under the condition that the prediction result is satisfactory or above, the energy consumption of the desk lamp is minimized when it is running at the current gear (the prediction results are either satisfactory or very satisfactory. For both prediction conditions, the desk lamp with the lowest energy consumption is selected to achieve the purpose of minimizing energy consumption); Feedback the optimization results to the execution unit to dynamically adjust the curtains and lamps. Figure 9As shown, based on the prediction results, the curtain opening and closing and lamp gear change requirements are made. If there are curtains and desk lamps on the market that can dynamically adjust the opening and closing degrees according to the control instructions, the intelligent lighting system can be introduced to generate command signals through the prediction results and transmit them to the control end to control the changes of curtains and lamps. If there is no ready-made technology on the market, a dynamic signal control system can be manually designed. For curtains, the voltage receiving end of the electric curtain motor is connected to components such as silicon modules. By changing the PWM signal to control the output voltage of the silicon module, the motor operating voltage is adjusted, and the curtain opening is dynamically adjusted. The same is true for desk lamps.

[0061] This embodiment dynamically collects light environment data, calculates indoor illumination gradients, predicts light satisfaction based on illumination gradients, and optimizes indoor curtain openings and lamp positions according to the predicted light satisfaction values, thereby achieving personalized and energy-efficient indoor lighting. This effectively solves the problem of lack of coordinated control between artificial lighting and natural light, and dynamically adjusts curtain openings and lamp positions based on light satisfaction, thereby improving the intelligence level of indoor lighting regulation. At the same time, through personalized adjustment combined with student satisfaction scores, a "one-person-one-policy" light environment optimization is achieved, visual comfort is improved, natural light changes are responded to in real time, and lighting strategies are dynamically adjusted through prediction and optimization algorithms.

[0062] Example 2 The second embodiment of the present invention introduces an indoor lighting control system based on illumination gradient.

[0063] like Figure 10 An indoor lighting control system based on illumination gradient is shown, comprising: A collection module configured to collect dynamic light environment data in a room; A calculation module configured to calculate an indoor illumination gradient based on light environment data; a prediction module configured to predict indoor light satisfaction based on the obtained illumination gradient; The control module is configured to optimize the combination of curtain opening and lamp level according to the light satisfaction prediction value, and dynamically adjust the curtains and lamps based on the combined optimization results to complete the indoor lighting control based on the illumination gradient.

[0064] The detailed steps are the same as those of the indoor lighting control method based on illumination gradient provided in Example 1, and will not be repeated here.

[0065] Example 3 A third embodiment of the present invention provides a computer-readable storage medium.

[0066] A computer-readable storage medium stores a program thereon, which, when executed by a processor, implements the steps of the indoor lighting control method based on illumination gradient as described in the first embodiment of the present invention.

[0067] The detailed steps are the same as those of the indoor lighting control method based on illumination gradient provided in Example 1, and will not be repeated here.

[0068] Example 4 A fourth embodiment of the present invention provides an electronic device.

[0069] An electronic device includes a memory, a processor, and a program stored in the memory and running on the processor. When the processor executes the program, the steps of the indoor lighting control method based on illumination gradient as described in Example 1 of the present invention are implemented.

[0070] The detailed steps are the same as those of the indoor lighting control method based on illumination gradient provided in Example 1, and will not be repeated here.

[0071] Example 5 A fifth embodiment of the present invention provides a computer program product.

[0072] A computer program product includes software code, wherein the program in the software code executes the steps of the indoor lighting control method based on illumination gradient as described in the first embodiment of the present invention.

[0073] The detailed steps are the same as those of the indoor lighting control method based on illumination gradient provided in Example 1, and will not be repeated here.

[0074] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk drives, CD-ROMs, optical storage devices, etc.) containing computer-usable program code. The solutions in the embodiments of the present invention may be implemented using various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0075] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0076] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0077] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0078] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0079] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

[0080] The above description is merely a preferred embodiment of this embodiment and is not intended to limit this embodiment. Those skilled in the art will readily appreciate that this embodiment may be modified and varied in various ways. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of this embodiment shall be within the scope of protection of this embodiment.

Claims

1. A method for controlling indoor lighting based on illumination gradient, characterized in that: include: Collect dynamic light environment data indoors; Calculate the indoor illumination gradient based on the light environment data; Based on the obtained illuminance gradient, predict the indoor light satisfaction; The curtain opening and lamp level are optimized according to the light satisfaction prediction value. The curtains and lamps are dynamically adjusted based on the combined optimization results to complete the indoor lighting control based on the illumination gradient.

2. The indoor lighting control method based on illumination gradient as claimed in claim 1, characterized in that: Before calculating the indoor light satisfaction, the interpolation method is used to process the collected dynamic light environment data to obtain the illuminance distribution map at different locations in the room; in the obtained illuminance distribution map, the illuminance value decreases with the increase of the distance from the indoor and outdoor windows.

3. The indoor lighting control method based on illumination gradient as claimed in claim 1, characterized in that: The illumination gradient is the ratio of the difference between the illumination at the center point of the indoor desktop and the illumination at each edge point to the distance between the center point and the edge point; on the indoor horizontal cross section, the illumination gradient in different directions is defined with the intersection of the cross section diagonals as the center and the diagonals and median line passing through the center as the directions.

4. The indoor lighting control method based on illumination gradient as claimed in claim 3, characterized in that: The illumination gradients in different directions are calculated to detect outliers. The detected outliers are eliminated by combining the interquartile range method. The illumination gradients in different directions after eliminating the outliers are analyzed, and the increase or decrease of illumination in that direction is determined based on the positive or negative value of the illumination gradient.

5. The indoor lighting control method based on illumination gradient as claimed in claim 1, characterized in that: A multi-layer perceptron model is used to predict light satisfaction. The multi-layer perceptron used includes an input layer, an output layer, and at least one hidden layer. It learns the nonlinear characteristics of the illumination gradient in a feedforward manner and completes the prediction of light satisfaction based on the classification and regression of the illumination gradient.

6. The indoor lighting control method based on illumination gradient as claimed in claim 1, characterized in that: In the process of optimizing the combination of curtain opening and lamp level, a multi-objective optimization algorithm is used to optimize the combination of curtain opening and lamp level. The curtain opening is roughly adjusted according to the light satisfaction prediction value until the indoor light satisfaction value is higher than the satisfaction threshold. The lamp level is adjusted in combination with the indoor satisfaction value, and the indoor artificial illumination value is changed to obtain the optimal combination of curtain opening and lamp level. The curtain opening and lamp level are dynamically adjusted in real time according to the obtained optimal combination.

7. An indoor lighting control system based on illumination gradient, characterized in that: include: A collection module configured to collect dynamic light environment data in a room; A calculation module configured to calculate an indoor illumination gradient based on light environment data; a prediction module configured to predict indoor light satisfaction based on the obtained illumination gradient; The control module is configured to optimize the combination of curtain opening and lamp level according to the light satisfaction prediction value, and dynamically adjust the curtains and lamps based on the combined optimization results to complete the indoor lighting control based on the illumination gradient.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the indoor lighting control method based on illumination gradient as described in any one of claims 1 to 6 are implemented.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the program, the steps of the indoor lighting control method based on illumination gradient are implemented as described in any one of claims 1 to 6.

10. A computer program product comprising software code, characterized in that The program in the software code executes the steps of the indoor lighting control method based on illumination gradient according to any one of claims 1 to 6.

Images