Public area light environment design system and method based on user comfort

Abstract

The invention relates to the field of light environment design methods, and particularly discloses a public area light environment design system and method for user comfort, and the method comprises the following steps: S1, obtaining the purpose information of a target public area, the purpose information comprising at least one of a commercial purpose, a traffic purpose and a leisure purpose; s2, according to the meteorological data, the geographic information and the building structure information of the target public area, the change conditions of the natural light incident angle, the illumination intensity and the illumination duration in the target public area are obtained; s3, according to the purpose of the target public area, simulating the illumination user flow density, the activity path and the stay area distribution, and arranging a model based on a pre-established sensor in combination with the natural light change condition; according to the technical scheme, personalized demand matching can be carried out on different functional areas, natural light can be fully utilized, and evaluation can be carried out after design is completed.

Description

Technical Field

[0001] This invention relates to the field of lighting environment design methods, and particularly to a system and method for designing public area lighting environments based on user comfort. Background Technology

[0002] In the process of urban construction and public space optimization, the importance of public area lighting design is becoming increasingly prominent. However, current mainstream lighting design technologies and methods still have many limitations that need to be addressed, making it difficult to meet the demands of modern society for high-quality public spaces.

[0003] From a design perspective, existing public area lighting designs mostly rely on national or industry-specific general lighting standards as their primary reference. For example, the "Standard for Lighting Design of Buildings" explicitly stipulates that the average illuminance in office spaces should reach 300-500 lux, and in shopping malls, it should reach 300 lux. However, this "one-size-fits-all" approach seriously ignores the diversity and uniqueness of public area uses. Taking commercial complexes as an example, supermarkets and bookstores, while both commercial entities, have drastically different lighting requirements: supermarkets need bright and uniform lighting to facilitate customer selection, while bookstores prioritize creating a soft and layered lighting atmosphere suitable for reading. Traditional design methods struggle to accurately adapt to the personalized needs of different functional areas within the same building, resulting in a disconnect between the lighting environment and actual usage scenarios, negatively impacting user experience.

[0004] There are also significant shortcomings in the utilization of natural light. Natural light, as a clean and comfortable light source, can effectively improve the quality of the lighting environment and user comfort in public areas. However, current designs often simply estimate the incident angle and intensity of natural light based on basic information such as window area and orientation, lacking in-depth research and dynamic simulation of the changing patterns of natural light with seasons, weather, and time. For example, in northern regions during winter, days are short and nights are long with weak light intensity. Traditional designs do not consider this change, potentially leading to a lack of effective integration and complementarity between artificial lighting and natural light, resulting in energy waste. In rainy southern regions, insufficient sunlight on cloudy days, coupled with the failure to adjust artificial lighting according to weather changes, can result in dim lighting in public areas, reducing user willingness to engage and comfort. Furthermore, existing designs rarely combine the changing characteristics of natural light with the functional needs of public areas, failing to fully leverage the potential of natural light in enhancing spatial atmosphere and user experience.

[0005] In the post-design evaluation phase, existing technologies also have shortcomings. Traditional design often relies solely on designers' subjective evaluations or simple tests of indicators such as illuminance and color temperature to judge the quality of lighting environment design, lacking simulation and evaluation of actual user experience. For example, when designing the nighttime lighting environment of a park, designers may only focus on whether the brightness of the lighting equipment meets standards, without considering changes in pedestrian density at different times of day or the differences in lighting needs of users in different scenarios (such as walking, resting, and exercising). This evaluation method, detached from actual usage scenarios, makes it impossible to discover and solve problems in the design in a timely manner, ultimately resulting in a public area lighting environment that fails to meet the diverse needs of users and cannot effectively improve user satisfaction and comfort.

[0006] Therefore, there is an urgent need for a public area lighting environment design system and method that can match the personalized needs of different functional areas, make full use of natural light, and be evaluated after the design is completed, based on user comfort. Summary of the Invention

[0007] This invention provides a public area lighting environment design system and method based on user comfort, which can match the personalized needs of different functional areas, make full use of natural light, and evaluate the design after completion.

[0008] To solve the above-mentioned technical problems, this application provides the following technical solution:

[0009] The public area lighting environment design method based on user comfort includes the following steps:

[0010] S1, Obtain the usage information of the target public area, wherein the usage information includes at least one of commercial use, transportation use, and leisure use;

[0011] S2, by using meteorological data, geographic information and building structure information of the target public area, obtains the changes in the natural light incident angle, light intensity and light duration in the target public area;

[0012] S3. Based on the purpose of the target public area, simulate the distribution of lighting users' flow density, activity paths, and dwelling areas. Combined with the changes in natural light, and based on the pre-established sensor layout model, automatically set the sensor collection points for collecting lighting user comfort-related data. The comfort-related data includes ambient temperature and humidity, light intensity, and dwell time.

[0013] S4. Virtually deploy sensors at the sensor acquisition points to simulate real-time data acquisition, and then input the data into a pre-developed light environment control model. The light environment control model automatically calculates the light environment design parameters based on the acquired data to obtain a preliminary design. The design parameters include the illuminance, color temperature, switching time, and illumination direction of the lighting equipment.

[0014] S5, collect the user's adjustment requirements for the lighting environment design, incorporate the adjustment requirements into the lighting environment control model, update the lighting environment design parameters, and adjust the preliminary design to obtain a revised design;

[0015] S6, Introduce simulated user flow in the target public area of ​​the modified design, collect user satisfaction data in the simulated scene through the virtual deployment sensors, and generate feedback optimization suggestions for the lighting environment design based on the satisfaction data;

[0016] S7. The feedback optimization suggestions are displayed to the user. After the user confirms, the lighting environment design of the target public area is completed according to the final determined lighting environment design parameters.

[0017] The basic principles and beneficial effects of the proposed solution are as follows: The core principle of the user comfort-based public area lighting environment design method described in this invention lies in achieving precise and personalized lighting environment design through multi-dimensional data collection, analysis, and dynamic simulation. First, the intended use of the target public area is obtained, clarifying the area's attributes (e.g., commercial, transportation, leisure). Different uses have fundamentally different lighting environment requirements. For example, commercial areas emphasize product display and require high illuminance and high color rendering index light, while leisure areas emphasize a softer, more comfortable atmosphere. This step lays the foundation for subsequent design.

[0018] Next, meteorological data, geographic information, and building structure information are used to obtain information on changes in natural light. As an important component of the lighting environment in public areas, the dynamic changes in natural light over time, season, and weather significantly affect user experience and lighting energy consumption. Through astronomical algorithms and ray tracing algorithms, the propagation and distribution of natural light inside buildings can be accurately simulated, thus providing a reference for artificial lighting design and achieving an organic combination of natural and artificial light.

[0019] Subsequently, user flow patterns were simulated based on the use of public areas, and data collection points were automatically set based on a pre-established sensor deployment model, taking into account changes in natural light. This process considered user activity patterns and dwell time in different areas to ensure that sensors could collect key data that best reflects user comfort. For example, more sensors were deployed in areas with high foot traffic and long dwell times to obtain accurate data.

[0020] After acquiring data, virtual sensors are deployed to simulate and input the collected data into the lighting environment control model, calculating preliminary design parameters. This model integrates multiple factors such as user comfort and regional functional requirements, and optimizes the output lighting environment parameters to meet expectations. Subsequently, user adjustments are collected to revise the preliminary design, demonstrating interactivity and dynamic adaptability in the design process. Finally, simulated user flow is used to obtain satisfaction data and generate optimization suggestions, which are then confirmed by the users to complete the design. This comprehensive approach, considering both objective environmental data and subjective user needs, ensures the scientific rigor and practicality of the lighting environment design.

[0021] Compared to traditional fixed-standard lighting environment design, this method, by clearly defining the purpose of public areas and simulating user flow, can precisely design lighting environments to meet the differentiated needs of users in different scenarios. For example, in public areas used for transportation, the brightness and distribution of lighting can be rationally planned based on passenger movement and waiting patterns, avoiding excessively strong or dim light that could negatively impact the user experience and greatly improving user comfort in different scenarios.

[0022] By accurately capturing and simulating variations in natural light, artificial lighting systems can dynamically coordinate with natural light. This reduces the use of artificial lighting during periods of ample natural light, lowering energy consumption; and supplements natural light promptly when it is insufficient, ensuring the quality of the lighting environment. Compared to traditional designs that ignore changes in natural light, this effectively improves energy efficiency and aligns with the trend of green and energy-saving development.

[0023] This model automatically sets the data collection points based on a pre-established sensor layout model, changing the traditional approach of relying on subjective experience to place sensors. It comprehensively considers user activities and natural light factors, ensuring the comprehensiveness and validity of the sensor data, providing a reliable basis for subsequent lighting environment design and optimization, and avoiding design flaws caused by missing or biased data.

[0024] The design process incorporates user feedback and adjustments based on user requests, which are then incorporated into the model for refinement. Furthermore, optimization is achieved through simulated user flow and satisfaction feedback, forming a dynamic interactive mechanism. Unlike traditional static design methods, this approach responds promptly to user needs, ensuring the final lighting environment design better aligns with actual user expectations and significantly enhancing the rationality and practicality of the design solution.

[0025] In summary, this invention achieves personalized matching of different functional areas, makes full use of natural light, and enables evaluation and optimization suggestions after the design is completed.

[0026] Further, S1 obtains the usage information of the target public area, specifically by: using a text analysis model to perform semantic analysis on architectural planning documents, lease contracts, and other materials related to the target public area; calculating the semantic similarity between keywords in the documents and commercial, transportation, and leisure uses using a word vector model; setting a similarity threshold of α (0 < α < 1); if the semantic similarity between a keyword and a certain use is greater than or equal to α, then the public area is determined to contain that use; wherein, the word vector model is trained on a corpus in the architectural field using the Word2Vec algorithm, and the semantic similarity calculation formula is: A and B are the word vectors corresponding to keywords and usage categories, respectively.

[0027] Furthermore, in step S2, the changes in natural light within the target public area are obtained, specifically by establishing a solar position calculation model; astronomical algorithms are used to calculate the solar declination δ, hour angle ω, and solar altitude angle h. ⊙ Sun azimuth A ⊙ The calculation formula is as follows: Solar declination: Where n is the number of days in a year; hour angle: ω = 15°·(t-12), t is the local time; solar altitude angle: sinh ⊙ =sinφ·sinδ+cosφ·cosδ·cosω, where φ is the geographical latitude of the target public area; solar azimuth angle: By combining the building structure information of the target public area, the reflection and refraction of natural light inside the building are simulated using ray tracing algorithms to obtain the changes in the incident angle, light intensity and duration of natural light in each area.

[0028] Furthermore, the sensor layout model pre-established in S3 is constructed as follows: First, the target common area is divided into N grid areas, and each grid area i corresponds to a weight value w. i The weight value is obtained through the formula Calculate, where f i For the projected user traffic in this area, t i S is the estimated average time users spend in the room. i Let H be the area of ​​the region; then, with the goal of maximizing the information entropy H of the collected data, the sensor placement is optimized using a genetic algorithm. The formula for calculating the information entropy is: p i The probability of collecting valid data in grid region i is given by ; the final sensor acquisition point is the center position of the grid region corresponding to the maximum information entropy after the genetic algorithm iteration.

[0029] Furthermore, the light environment control model is constructed based on a multi-objective optimization algorithm, with user comfort, energy efficiency, and lighting uniformity as optimization objectives; the user comfort is quantified by a comfort rating function C, and the calculation formula is as follows: Where I is the current light intensity, I min and I max The lower and upper limits of suitable light intensity for the target public area for its intended use, where T is the current color temperature. min and T max For the appropriate lower and upper limits of color temperature for the corresponding application, β1 and β2 are weighting coefficients and β1 + β2 = 1; the energy efficiency is represented by energy consumption per unit area E. Where P j Let S be the power of the j-th lighting device. j Let S be the coverage area of ​​the j-th lighting device. total The target public area is the total area; the lighting uniformity is calculated using the uniformity index U. I min,area I represents the minimum illumination intensity for the target area. avg,area The average light intensity of the target area is given; the multi-objective optimization problem is solved by the Non-dominated Sorting Genetic Algorithm-II (NSGA-II) to obtain the preliminary design parameters of the light environment.

[0030] Furthermore, the simulated real-time acquired data undergoes data preprocessing before being input into the light environment control model; a Kalman filter algorithm is used to denoise the simulated acquired data such as light intensity, ambient temperature, and humidity. For the light intensity data X... k The prediction and update formulas are as follows: Prediction equation: Update equation: in, For the prior estimate at time k, For the posterior estimate at time k-1, F k Let B be the state transition matrix. k To control the input matrix, u k K is the control vector. k For Kalman gain, Z k H is the observation value at time k. k The observation matrix is ​​used; the denoised data is then subjected to dimensionality reduction through principal component analysis (PCA) to extract key features before being input into the light environment control model, thereby improving the model's computational efficiency and the accuracy of design parameter calculations.

[0031] in, It is the core variable in the optical environment design parameter update model built based on the Kalman filter idea. Specifically, it represents the optimal estimate of the optical environment design parameters at time k, after combining the currently collected user adjustment demand data (i.e., observation values). k represents the time of iterative update (such as the first adjustment, the second adjustment, etc.). This represents the estimated value of the design parameters (distinguished from the true value X); k|k represents the estimate obtained at time k using the observation data at time k, to distinguish between them. At time k, the predicted value is obtained using only the data from time k-1 and earlier.

[0032] By integrating predicted values ​​of historical design parameters with real-time adjustment needs of current users, the lighting environment parameters (such as illuminance, color temperature, color rendering index, etc.) are dynamically optimized, making the revised design more in line with the actual needs of users.

[0033] Furthermore, the step in S5 of collecting the design user's adjustment requirements for the lighting environment design and incorporating these requirements into the lighting environment control model specifically includes: establishing a quantitative model of user adjustment requirements, and converting the adjustment requirements input by the design user into a computable parameter vector. Where r1 is the illuminance adjustment coefficient, r2 is the color temperature adjustment coefficient, r3 is the color rendering index adjustment coefficient, and r4 is the light and shadow distribution adjustment coefficient; the weight vector of each adjustment coefficient is determined by the Analytic Hierarchy Process (AHP). The weighting coefficients are obtained by constructing a judgment matrix and performing a consistency check; a demand response function is then constructed. The formula used to calculate the impact of adjustment requirements on lighting environment design parameters is: Where f i (r i Let be the influence function of the i-th adjustment coefficient, using the sigmoid activation function form: For shape parameters, θ i The threshold parameters are dynamically adjusted based on the intended use of the target public area. The output of the demand response function is used as a correction factor to adjust the preliminary design of the lighting environment parameters. The correction formula is as follows: Where P old For preliminary design parameters, P new The corrected design parameters are used to obtain the corrected design through this correction mechanism.

[0034] In this context, w1, w2, w3, and w4 are the weight values ​​of the adjustment coefficients determined in the Analytic Hierarchy Process (AHP), corresponding to the four key adjustment dimensions in lighting environment design. w1 corresponds to the illuminance adjustment coefficient, representing the importance of the user's adjustment needs for lighting environment brightness (such as average illuminance and minimum illuminance) in the overall optimization. For example, w1 may have a higher weight in an office area because illuminance directly affects work efficiency. w2 corresponds to the color temperature adjustment coefficient, representing the weight of the user's adjustment needs for warm or cool light (such as 3000K warm light and 6500K cool light). For example, w2 may be higher in a leisure area because color temperature directly affects the atmosphere. w3 corresponds to the color rendering index adjustment coefficient, representing the weight of the user's adjustment needs for the ability of light to reproduce the true colors of objects (such as Ra value). For example, w3 has a higher weight in a commercial display area because it is necessary to accurately present the colors of goods. w4 corresponds to the dynamic response speed adjustment coefficient, representing the weight of the user's adjustment needs for the speed at which lighting environment parameters respond to changes in the scene (such as the brightness adjustment speed when there is a sudden change in pedestrian flow). For example, w4 may have a higher weight in transportation hubs because it needs to adapt quickly to changes in passenger flow.

[0035] The weight vector W = (w1, w2, w3, w4) must satisfy w1 + w2 + w3 + w4 = 1, and its rationality is ensured by constructing a judgment matrix and a consistency check (CR < 0.1). Finally, it is used to weight and integrate the adjustment requirements of each dimension to generate comprehensive optimization suggestions.

[0036] Furthermore, the process of constructing simulated user flow introduced in S6 specifically involves: establishing a user behavior characteristic model based on the usage information of the target public area, dividing the simulated users into n behavioral groups, with each group corresponding to a different activity mode weight vector. Where w i1 As the weight of dwell time behavior, w i2 For the weight of the movement behavior, w i3 As the weight of the interaction behavior, and Using the Monte Carlo simulation algorithm, the behavior sequence of each simulated user within a unit of time is randomly generated based on the weight vector. The formula for calculating the behavior sequence is as follows: Here, rand() is a function that generates random numbers between 0 and 1, and t is the simulation duration. The generated behavioral sequence is mapped onto a three-dimensional spatial model of the target public area to form a dynamically changing simulated user flow, which is used to simulate the interaction between users and the light environment in a real scene.

[0037] Furthermore, the process of generating feedback optimization suggestions based on satisfaction data in S6 specifically involves: constructing a user satisfaction evaluation neural network model, which includes an input layer, a hidden layer, and an output layer. The input layer receives environmental parameter data (light intensity, color temperature, ambient temperature and humidity, etc.) collected by virtual deployment sensors and simulated user behavior data (stay time, movement speed, behavior type, etc.); in the hidden layer, a custom activation function is used... (Where a and b are parameters adaptively adjusted based on training data) Feature extraction and nonlinear transformation are performed on the input data; the output layer outputs the user's satisfaction rating S for the lighting environment, with a rating range of [0,1]; the backpropagation algorithm is used to train the neural network model, and the loss function is... Where S pred,i S represents the satisfaction score predicted by the model. true,i To simulate real user satisfaction feedback, the collected satisfaction data was analyzed based on a trained model. A sensitivity analysis algorithm was used to calculate the influence coefficient η of each lighting environment design parameter on the satisfaction score. j Based on the magnitude of the influence coefficient, according to the formula R j =sgn(η j )·|η j |·ΔP j (where sgn is the sign function, ΔP) j (Adjust the step size for design parameters) Generate feedback optimization suggestions for each design parameter. Attached Figure Description

[0038] Figure 1 A flowchart illustrating a user-comfort-based approach to designing public area lighting environments. Detailed Implementation

[0039] The following detailed description illustrates the specific implementation method:

[0040] Public area lighting environment design methods based on user comfort (e.g.) Figure 1 (As shown), including the following steps:

[0041] S1, Obtain the usage information of the target public area, wherein the usage information includes at least one of commercial use, transportation use, and leisure use;

[0042] S2, by using meteorological data, geographic information and building structure information of the target public area, obtains the changes in the natural light incident angle, light intensity and light duration in the target public area;

[0043] S3. Based on the purpose of the target public area, simulate the distribution of lighting users' flow density, activity paths, and dwelling areas. Combined with the changes in natural light, and based on the pre-established sensor layout model, automatically set the sensor collection points for collecting lighting user comfort-related data. The comfort-related data includes ambient temperature and humidity, light intensity, and dwell time.

[0044] S4. Virtually deploy sensors at the sensor acquisition points to simulate real-time data acquisition, and then input the data into a pre-developed light environment control model. The light environment control model automatically calculates the light environment design parameters based on the acquired data to obtain a preliminary design. The design parameters include the illuminance, color temperature, switching time, and illumination direction of the lighting equipment.

[0045] S5, collect the user's adjustment requirements for the lighting environment design, incorporate the adjustment requirements into the lighting environment control model, update the lighting environment design parameters, and adjust the preliminary design to obtain a revised design;

[0046] S6, Introduce simulated user flow in the target public area of ​​the modified design, collect user satisfaction data in the simulated scene through the virtual deployment sensors, and generate feedback optimization suggestions for the lighting environment design based on the satisfaction data;

[0047] S7. The feedback optimization suggestions are displayed to the user. After the user confirms, the lighting environment design of the target public area is completed according to the final determined lighting environment design parameters.

[0048] Specifically, S1 obtains the usage information of the target public area by: using a text analysis model to perform semantic analysis on architectural planning documents, lease contracts, and other materials related to the target public area; calculating the semantic similarity between keywords in the documents and commercial, transportation, and leisure uses using a word vector model; setting a similarity threshold of α (0 < α < 1); if the semantic similarity between a keyword and a certain use is greater than or equal to α, then the public area is determined to contain that use; wherein, the word vector model is trained on a corpus in the architectural field using the Word2Vec algorithm, and the semantic similarity calculation formula is: A and B are the word vectors corresponding to keywords and usage categories, respectively.

[0049] Taking the lighting environment design of the waiting hall of a large integrated transportation hub as an example, the first step is to obtain the usage information of the target public area, namely the waiting hall. Then, a text analysis model is used to perform semantic parsing on the hub's architectural planning documents, lease agreements, and other materials. A word vector model trained on an architectural corpus using the Word2Vec algorithm is employed to calculate the semantic similarity between keywords in the documents and their commercial, transportation, and leisure uses.

[0050] Assuming the document contains keywords such as "waiting," "ticket check," and "passenger passage," the semantic similarity between "waiting" and transportation use is calculated to be 0.85, and between "ticket check" and transportation use to be 0.9, both exceeding the set similarity threshold α = 0.8. Therefore, the waiting hall is determined to have transportation use. Simultaneously, the document contains keywords such as "convenience store" and "coffee shop." The semantic similarity between "convenience store" and commercial use is 0.88, and between "coffee shop" and commercial use to be 0.86, also exceeding the threshold α. Therefore, the waiting hall also has commercial use. Ultimately, the intended use of this target public area is determined to be both transportation and commercial.

[0051] In S2, the changes in natural light within the target public area are obtained, specifically by establishing a solar position calculation model; astronomical algorithms are used to calculate the solar declination δ, hour angle ω, and solar altitude angle h. ⊙ Sun azimuth A ⊙ The calculation formula is as follows: Solar declination: Where n is the number of days in a year; hour angle: ω = 15°·(t-12), t is the local time; solar altitude angle: The geographical latitude of the target public area; solar azimuth: By combining the building structure information of the target public area, the reflection and refraction of natural light inside the building are simulated using ray tracing algorithms to obtain the changes in the incident angle, light intensity and duration of natural light in each area.

[0052] By establishing a model for calculating the sun's position, astronomical algorithms are used to calculate the sun's declination δ, hour angle ω, and solar altitude angle h. ⊙ Sun azimuth A ⊙ The waiting hall is located at 30° North latitude. Assuming the calculation date is the 120th day of the year, then: solar declination, (in radians), the hour angle at 10:00 local time ω=15°·(10-12)=-30° (converted to approximately -0.524 radians), and the solar altitude angle sinh ⊙ =sin30°·sin0.139+cos30°·cos0.139·cos(-0.524)≈0.766, then h ⊙ ≈arcsin(0.766)≈0.876 (radians, approximately 50.2°), the solar azimuth angle. Then A ⊙ ≈arccos(-0.643)≈2.27 (radians, approximately 130°).

[0053] Based on the architectural structure information of the waiting hall, a ray tracing algorithm was used to simulate the reflection and refraction of natural light inside the building. Assuming the waiting hall has multiple large glass curtain walls, the simulation showed that from 8:00 AM to 10:00 AM, the natural light intensity in the eastern area of ​​the hall reached 500 lux, while the western area, due to obstruction, only had 100 lux. As time progressed, at noon, the light intensity in the central part of the hall reached a peak of 800 lux, while the edge areas reached 400 lux. This allowed for the acquisition of changes in the incident angle, intensity, and duration of natural light in each area at different times.

[0054] The pre-established sensor layout model in S3 is constructed as follows: First, the target common area is divided into N grid areas, and each grid area i corresponds to a weight value w. i The weight value is obtained through the formula Calculate, where f i For the projected user traffic in this area, t i S is the estimated average time users spend in the room. i Let H be the area of ​​the region; then, with the goal of maximizing the information entropy H of the collected data, the sensor placement is optimized using a genetic algorithm. The formula for calculating the information entropy is: p i The probability of collecting valid data in grid region i is given by ; the final sensor acquisition point is the center position of the grid region corresponding to the maximum information entropy after the genetic algorithm iteration.

[0055] The lighting environment control model is constructed based on a multi-objective optimization algorithm, with user comfort, energy efficiency, and lighting uniformity as optimization objectives. User comfort is quantified using a comfort rating function C, calculated as follows: Where I is the current light intensity, I min and I max The lower and upper limits of suitable light intensity for the target public area for its intended use, where T is the current color temperature. min and T max For the appropriate lower and upper limits of color temperature for the corresponding application, β1 and β2 are weighting coefficients and β1 + β2 = 1; the energy efficiency is represented by energy consumption per unit area E. Where P j Let S be the power of the j-th lighting device. j Let S be the coverage area of ​​the j-th lighting device. total The target public area is the total area; the lighting uniformity is calculated using the uniformity index U. I min,area I represents the minimum illumination intensity for the target area. avg,areaThe average light intensity of the target area is given; the multi-objective optimization problem is solved by the Non-dominated Sorting Genetic Algorithm-II (NSGA-II) to obtain the preliminary design parameters of the light environment.

[0056] Before inputting the simulated real-time acquired data into the light environment control model, data preprocessing is performed. A Kalman filter algorithm is used to denoise the simulated acquired data on light intensity, ambient temperature, and humidity. For the light intensity data X... k The prediction and update formulas are as follows: Prediction equation: Update equation: in, For the prior estimate at time k, For the posterior estimate at time k-1, F k Let B be the state transition matrix. k To control the input matrix, u k K is the control vector. k For Kalman gain, Z k H is the observation value at time k. k The observation matrix is ​​used; the denoised data is then subjected to dimensionality reduction through principal component analysis (PCA) to extract key features before being input into the light environment control model, thereby improving the model's computational efficiency and the accuracy of design parameter calculations.

[0057] Based on the transportation and commercial uses of the waiting hall, the simulation study analyzed the pedestrian density, activity paths, and dwell area distribution using lighting. The waiting hall was divided into 100 grid areas, each with an area of ​​10 square meters. Analysis showed that the expected pedestrian flow in the concentrated seating area was f1 = 50 people / hour, and the expected average dwell time was t1 = 30 minutes (0.5 hours). Therefore, the weight value for this area was determined. The estimated foot traffic around the convenience store is f2 = 30 people / hour, and the estimated average dwell time is t2 = 10 minutes (1 / 6 hour). What is the weight value of this area? With the goal of maximizing the information entropy H of the collected data, a genetic algorithm was used to optimize the sensor placement. The initial population size was set at 50 individuals. After 50 generations of iteration, it was finally determined that the sensor collection points, which maximized the information entropy, were located at the center of key grid areas such as the center of the waiting area with concentrated seating, the entrance of the convenience store, and the vicinity of the ticket gate. A total of 10 sensors were deployed to collect comfort-related data such as ambient temperature and humidity, light intensity, and the duration of people's stay.

[0058] The step in S5, which involves collecting the user's adjustment requirements for the lighting environment design and incorporating these requirements into the lighting environment control model, specifically includes: establishing a quantitative model of user adjustment requirements, and converting the user's input adjustment requirements into a computable parameter vector. Where r1 is the illuminance adjustment coefficient, r2 is the color temperature adjustment coefficient, r3 is the color rendering index adjustment coefficient, and r4 is the light and shadow distribution adjustment coefficient; the weight vector of each adjustment coefficient is determined by the Analytic Hierarchy Process (AHP). The weighting coefficients are obtained by constructing a judgment matrix and performing a consistency check; a demand response function is then constructed. The formula used to calculate the impact of adjustment requirements on lighting environment design parameters is: Where f i (r i Let be the influence function of the i-th adjustment coefficient, using the sigmoid activation function form: For shape parameters, θ i The threshold parameters are dynamically adjusted based on the intended use of the target public area. The output of the demand response function is used as a correction factor to adjust the preliminary design of the lighting environment parameters. The correction formula is as follows: Where P old For preliminary design parameters, P new The corrected design parameters are used to obtain the corrected design through this correction mechanism.

[0059] The construction process of the simulated user flow introduced in S6 is as follows: Based on the usage information of the target public area, a user behavior feature model is established, and the simulated users are divided into n behavioral groups, each group corresponding to a different activity mode weight vector. Where w i1 As the weight of dwell time behavior, w i2 For the weight of the movement behavior, w i3 As the weight of the interaction behavior, and Using the Monte Carlo simulation algorithm, the behavior sequence of each simulated user within a unit of time is randomly generated based on the weight vector. The formula for calculating the behavior sequence is as follows: Here, rand() is a function that generates random numbers between 0 and 1, and t is the simulation duration. The generated behavioral sequence is mapped onto a three-dimensional spatial model of the target public area to form a dynamically changing simulated user flow, which is used to simulate the interaction between users and the light environment in a real scene.

[0060] Sensors are virtually deployed at sensor acquisition points to simulate real-time data collection, which is then input into a pre-developed light environment control model. The light environment control model is constructed based on a multi-objective optimization algorithm, with user comfort, energy efficiency, and lighting uniformity as optimization objectives. User comfort is quantified using a comfort rating function C. For the waiting hall, β1 = 0.6, β2 = 0.4, and a suitable lower limit of illumination intensity I are set. min =300 lux, upper limit I max =500 lux, suitable lower limit of color temperature T min=3000K, upper limit T max =4000K. Assuming a light intensity of I = 350 lux and a color temperature of T = 3500K is collected at a certain moment, then the comfort score... Energy efficiency is represented by energy consumption per unit area, E. Assuming the total area of ​​the waiting hall is S... total =1000 square meters, with a total of 50 lighting fixtures, of which 20 have a power of 50 watts and 30 have a power of 30 watts. The coverage area of ​​each fixture is determined based on the actual layout. The energy consumption per unit area is calculated. (Assuming E = 15 watts / square meter after calculation). Illumination uniformity is calculated using the uniformity index U, assuming a minimum illuminance I in a certain area. min,area =320 lux, average illuminance I avg,area =380 lux, then The multi-objective optimization problem was solved using the Non-Dominated Sorting Genetic Algorithm-II (NSGA-II), and the preliminary design parameters for the lighting environment were obtained: the illuminance was set at 380 lux, the color temperature was set at 3600K, the switching time of some lighting equipment was dynamically adjusted according to the natural light intensity, and the illumination direction was calculated and adjusted to make the light cover the waiting area more evenly.

[0061] The system collects user feedback on lighting environment design adjustments and incorporates these adjustments into the lighting environment control model. A quantitative model of user adjustment requirements is established, converting the user-input adjustments into a computable parameter vector. Suppose the user requests increased light intensity. Analysis determines the following adjustment coefficients: illuminance adjustment coefficient r1 = 0.1, color temperature adjustment coefficient r2 = 0, color rendering index adjustment coefficient r3 = 0, and light and shadow distribution adjustment coefficient r4 = 0. The weight vectors for each adjustment coefficient are then determined using the Analytic Hierarchy Process (AHP). Assuming that after constructing the judgment matrix and performing a consistency check, we obtain w1 = 0.7, w2 = 0.1, w3 = 0.1, and w4 = 0.1, then construct the demand response function. Using the sigmoid activation function, with k1 = 10 and θ1 = 0, then Demand response function The output of the demand response function is used as a correction factor to adjust the initial design of the lighting environment parameters, including the original illuminance P. old =380 lux, then the corrected illuminance P new =380·(1+0.512)=574.56 lux. Based on the actual situation, it is adjusted to 570 lux, thus completing the revision of the preliminary design and obtaining the revised design.

[0062] The process of generating feedback optimization suggestions based on satisfaction data in S6 is as follows: A neural network model for user satisfaction evaluation is constructed. This model includes an input layer, a hidden layer, and an output layer. The input layer receives environmental parameter data (light intensity, color temperature, ambient temperature and humidity, etc.) collected by virtual deployed sensors and simulated user behavior data (stay time, movement speed, behavior type, etc.). In the hidden layer, a custom activation function is used... (Where a and b are parameters adaptively adjusted based on training data) Feature extraction and nonlinear transformation are performed on the input data; the output layer outputs the user's satisfaction rating S for the lighting environment, with a rating range of [0,1]; the backpropagation algorithm is used to train the neural network model, and the loss function is... Where S pred,i S represents the satisfaction score predicted by the model. true,i To simulate real user satisfaction feedback, the collected satisfaction data was analyzed based on a trained model. A sensitivity analysis algorithm was used to calculate the influence coefficient η of each lighting environment design parameter on the satisfaction score. j Based on the magnitude of the influence coefficient, according to the formula R j =sgn(η j )·|η j |·ΔP j (where sgn is the sign function, ΔP) j (Adjust the step size for design parameters) Generate feedback optimization suggestions for each design parameter.

[0063] Simulated user flow was introduced into the redesigned waiting hall. Based on the traffic and commercial use information of the waiting hall, a user behavior characteristic model was established, dividing simulated users into three behavioral groups: waiting group, shopping group, and short-term passerby group. The activity pattern weight vector of the waiting group was then calculated. Shopping group activity pattern weight vector Activity pattern weight vector of short-term passersby

[0064] Using the Monte Carlo simulation algorithm, assuming a simulation duration of t = 60 minutes, the behavioral sequence of each simulated user within a unit of time is generated. For example, for a simulated waiting user, the behavioral sequence is... The user's dwell time, movement, and interaction behavior at different times are calculated and mapped onto a 3D spatial model of the waiting hall, forming a dynamically changing simulated user flow. Satisfaction data of simulated users in the simulated scenario is collected through virtually deployed sensors, and feedback optimization suggestions are generated based on this data. A neural network model for user satisfaction evaluation is constructed, which includes an input layer (receiving data such as light intensity, color temperature, ambient temperature and humidity, simulated user dwell time, movement speed, and behavior type) and a hidden layer (using a custom activation function). The input data undergoes feature extraction and nonlinear transformation, and the output layer outputs the user's satisfaction rating of the lighting environment (S).

[0065] The neural network model is trained using the backpropagation algorithm. Assume the training data contains 1000 sets of actual collected data, and the loss function is... After training, the newly collected satisfaction data is analyzed, and the influence coefficient η of each lighting environment design parameter on the satisfaction score is calculated using a sensitivity analysis algorithm. j Assuming the calculated influence coefficients for illuminance η1 = 0.6 and color temperature η2 = 0.3, then according to the formula R... j =sgn(η j )·|η j |·ΔP j (Setting ΔP1 = 20 lux, ΔP2 = 100K), the generated feedback optimization suggestion for illuminance is to increase by 12 lux, and the feedback optimization suggestion for color temperature is to increase by 30K.

[0066] The generated feedback and optimization suggestions are displayed to the user. After the user confirms, the lighting environment design for the target public area, namely the waiting hall, is completed based on the final determined lighting environment design parameters, such as adjusting the illuminance to 582 lux and the color temperature to 3630K.

[0067] The above are merely embodiments of the present invention. The invention is not limited to the fields covered by these embodiments. Commonly known structures and characteristics in the solutions are not described in detail here. Those skilled in the art are aware of all common technical knowledge in the field prior to the application date or priority date, are able to access all existing technologies in that field, and have the ability to apply conventional experimental methods prior to that date. Those skilled in the art can, under the guidance of this application, improve and implement this solution in combination with their own capabilities. Some typical known structures or methods should not be obstacles for those skilled in the art to implement this application. It should be noted that those skilled in the art can make several modifications and improvements without departing from the structure of the present invention. These should also be considered within the scope of protection of the present invention, and will not affect the effectiveness of the implementation of the present invention or the practicality of the patent. The scope of protection claimed in this application should be determined by the content of its claims, and the specific embodiments described in the specification can be used to interpret the content of the claims.

Claims

1. A method for designing a light environment in a public area based on user comfort, characterized in that, The method comprises the following steps: S1, obtaining the use information of the target public area, the use information comprising at least one of commercial use, traffic use and leisure use; S2, obtaining the variation of the natural light incidence angle, the light intensity and the light duration in the target public area through meteorological data, geographic information and the building structure information of the target public area; S3, simulating the user flow density, the activity path and the stay area distribution of the lighting user according to the use of the target public area, combining the variation of the natural light, and automatically setting the sensor collection points for collecting the comfort-related data of the lighting user based on a pre-established sensor arrangement model, the comfort-related data comprising the environmental temperature and humidity, the light intensity and the personnel stay duration; S4, virtually deploying sensors at the sensor collection points to simulate real-time collection data, and then inputting the pre-developed light environment control model to automatically complete the calculation of the light environment design parameters to obtain a preliminary design according to the collected data; the design parameters comprising the illumination, the color temperature, the switching time and the irradiation direction of the lighting device; S5, collecting the adjustment requirements of the design user for the light environment design, and incorporating the adjustment requirements into the light environment control model to update the light environment design parameters and adjust the preliminary design to obtain a revised design; S6, introducing the simulated user flow into the target public area of the revised design, collecting the satisfaction data of the user in the simulated scene through the virtually deployed sensors, and generating feedback optimization suggestions for the light environment design based on the satisfaction data; S7, displaying the feedback optimization suggestions to the user, and completing the light environment design of the target public area according to the finally determined light environment design parameters after the user confirms.

2. The method for designing a light environment of a public area based on user comfort according to claim 1, characterized in that, The S1 obtains the use information of the target public area, which is to use a text analysis model to perform semantic analysis on the building planning documents and rental contracts of the target public area, calculate the semantic similarity of the keywords in the documents with the commercial use, the traffic use and the leisure use through a word vector model, set the similarity threshold as α, wherein 0 < α < 1, and if the semantic similarity of the keywords with a use is greater than or equal to α, it is determined that the public area contains this use; Wherein, the word vector model is trained by Word2Vec algorithm on the corpus of the building field, and the semantic similarity calculation formula is: A and B are the word vectors corresponding to the keywords and the use categories, respectively.

3. The method for designing a light environment of a public area based on user comfort according to claim 2, characterized in that, The variation of the natural light in the target public area in the S2 is obtained by establishing a solar position calculation model; The astronomical algorithm is used to calculate the solar declination δ, the hour angle ω and the solar altitude angle h ⊙ , the solar azimuth angle A ⊙ , and the calculation formula is as follows: Sun declination: Wherein n is the number of days in a year; Hour angle: ω = 15°·(t-12), t is the local time; Solar altitude angle: sinh ⊙ = sinφ·sinδ+ cosφ·cosδ·cosω, φ is the geographical latitude of the target public area; Solar azimuth: Combined with the building structure information of the target public area, the reflection and refraction of the natural light inside the building are simulated through a ray tracing algorithm to obtain the variation of the natural light incidence angle, the light intensity and the light duration in each area.

4. The method for designing a light environment of a public area based on user comfort according to claim 3, characterized in that, The construction process of the pre-established sensor arrangement model in the S3 is as follows: First, the target public area is divided into N grid areas, each grid area i corresponds to a weight value w i , which is calculated by the formula , where f i is the expected user traffic of the area, t i is the expected average user stay time, and S i is the area size. Then, the sensor placement position is optimized by genetic algorithm to maximize the information entropy H of the collected data, and the information entropy calculation formula is p i is the probability of collecting effective data in the grid area i; and the final sensor collection point is the center position of the grid area corresponding to the maximum information entropy after genetic algorithm iteration.

5. The method for designing a light environment of a public area based on user comfort according to claim 4, characterized in that, The light environment control model is constructed based on a multi-objective optimization algorithm, and the user comfort, the energy consumption efficiency and the lighting uniformity are taken as the optimization objectives; The user comfort is quantified by a comfort score function C, calculated as: where I is the current light intensity, I min and I max are the lower and upper limits of the suitable light intensity for the application of the target common area, T is the current color temperature, T min and T max are the lower and upper limits of the suitable color temperature for the application, and β1 and β2 are weight coefficients and β1 + β2 = 1. The energy consumption efficiency is expressed by the energy consumption per unit area E, where P j is the power of the jth lighting device, S j is the coverage area of the jth lighting device, S total is the total area of the target public area; The illumination uniformity is calculated by a uniformity index U, I min,area is the minimum illumination intensity of the target region, I avg,area is the average illumination intensity of the target region; the multi-objective optimization problem is solved by a non-dominated sorting genetic algorithm-II, i.e., NSGA-II, to obtain the preliminary designed light environment design parameters.

6. The method for designing a light environment of a public area based on user comfort according to claim 5, characterized in that, The simulated real-time collection data is preprocessed before being input into the light environment control model; The Kalman filtering algorithm is used to denoise the simulated and collected data of light intensity, environmental temperature and humidity, etc. The prediction and update formula of the light intensity data X k are as follows: Prediction equation: Update equation: wherein, is a prior estimate at time k, is a posterior estimate at time k-1, F k is a state transition matrix, B k is a control input matrix, u k is a control vector, K k is a Kalman gain, Z k is an observation value at time k, H k is an observation matrix; After denoising, the data is processed by principal component analysis for dimension reduction, and the key features are extracted before being input into the light environment control model, so as to improve the calculation efficiency of the model and the accuracy of the design parameter calculation.

7. The method for designing a light environment of a public area based on user comfort according to claim 6, characterized in that, The step of collecting the adjustment requirements of the design user on the light environment design in S5 and incorporating the adjustment requirements into the light environment control model comprises: A user adjustment demand quantification model is established to convert the adjustment demand input by the design user into a calculable parameter vector Wherein r1 is an illumination adjustment coefficient, r2 is a color temperature adjustment coefficient, r3 is a color rendering index adjustment coefficient, and r4 is a light and shadow distribution adjustment coefficient. The weight vector of each adjustment coefficient is determined by analytic hierarchy process The weight coefficient is obtained by constructing a judgment matrix and performing consistency check. Constructing a demand response function For calculating the degree of influence of adjusting demand on the light environment design parameters, the formula is: where f i (r i ) is the influence function of the ith adjustment coefficient, in the form of a sigmoid activation function: wherein k i is a shape parameter, θ i is a threshold parameter, both of which are dynamically adjusted according to the use category of the target common area; The output result of the demand response function is used as a correction factor to correct the light environment parameters of the preliminary design, and the correction formula is: where P old is the preliminary design parameter, P new is the corrected design parameter, and the corrected design is obtained through the correction mechanism.

8. The method for designing a light environment of a public area based on user comfort according to claim 7, characterized in that, The construction process of introducing the simulated user flow in S6 is: Based on the target public area use information, a user behavior feature model is established, and the simulation users are divided into n behavior groups, each group corresponding to a different activity mode weight vector wherein w i1 is a stay behavior weight, w i2 is a moving behavior weight, w i3 is an interaction behavior weight, and The behavior sequence of each simulation user in unit time is randomly generated according to the weight vector by using a Monte Carlo simulation algorithm, and the calculation formula of the behavior sequence is where rand() is a function of generating a random number between 0 and 1, and t is the simulation duration. The generated behavior sequence is mapped to the three-dimensional space model of the target public area to form a dynamically changing simulated user flow for simulating the interaction between the user and the light environment in a real scene.

9. The method for designing a light environment of a public area based on user comfort according to claim 8, characterized in that, The process of generating feedback optimization suggestions based on the satisfaction data in S6 is: A user satisfaction evaluation neural network model is constructed, which includes an input layer, a hidden layer and an output layer. The input layer receives environmental parameter data collected by virtual deployment sensors and behavior data of simulated users; In the hidden layer, through the self-defined activation function where a, b are parameters that are adaptively adjusted according to the training data, and the input data is subjected to feature extraction and nonlinear transformation; The output layer outputs the user's satisfaction score S for the light environment, and the score range is [0, 1]; The neural network model is trained by using a back propagation algorithm, and a loss function is where S pred,i is a satisfaction score predicted by the model, and S true,i is a real satisfaction feedback of the simulated user. Based on the trained model, the collected satisfaction data is analyzed, and the influence coefficient η of each light environment design parameter on the satisfaction score is calculated through a sensitivity analysis algorithm j , according to the size of the influence coefficient, according to the formula R j = sgn(η j )·|η j |·ΔP j , wherein sgn is a sign function, ΔP j is the design parameter adjustment step, and feedback optimization suggestions for each design parameter are generated.

10. A public area light environment design system based on user comfort, characterized by, The method of any one of claims 1-9 is adopted.

Images