Energy-saving illumination management system based on three-dimensional visualization technology

Abstract

The invention discloses an energy-saving illumination management system based on a three-dimensional visualization technology, and belongs to the field of energy-saving illumination management systems.The system makes an accurate optimal illumination strategy based on multi-source data decision, meanwhile, a three-dimensional scene construction module can adjust the illumination strategy in real time according to actual changes in an area, and the illumination strategy is optimized. According to the method, illumination is ensured to meet actual requirements all the time, the introduced simulated annealing algorithm can find a global optimal solution in numerous possible illumination strategies, and the introduced simulated annealing algorithm accepts a new strategy according to a Metropolis criterion in an iteration process by setting parameters such as an initial temperature, a temperature drop rate and an end temperature, so that the optimal illumination strategy is obtained. The method can effectively cope with changes of use properties in different time periods and different illumination areas, flexibly adjust illumination strategies in different scenes, comprehensively consider energy consumption and illumination demand matching degree in cost function calculation, calculate illumination matching degree in combination with illumination sensor data, balance weights of the two by adjusting coefficients, and improve the accuracy of cost function calculation. And the illumination quality is ensured not to be sacrificed while energy conservation is pursued.

Description

Technical Field

[0001] The present invention relates to the field of energy-saving lighting management systems, and more specifically, to an energy-saving lighting management system based on three-dimensional visualization technology. Background Art

[0002] In today's society, lighting systems have been deeply integrated into people's production and life. Whether it is an office, a commercial area or a living environment, lighting is an indispensable part. However, there are many drawbacks in traditional lighting management systems, resulting in widespread energy waste.

[0003] For example, in an office environment, most offices adopt a unified lighting control method, and the lights are turned on and off at fixed times. However, in actual work, it often happens that employees work overtime, come to work early or move around in different areas. For example, when only a few employees work overtime in a certain office area after work, all the lights in the entire office area are still turned on and the brightness is fixed, which undoubtedly causes a large amount of electrical energy waste. Moreover, traditional lighting systems are not very convenient to adjust in real time according to the change of indoor natural light intensity. When the sun is sufficient during the day, if the indoor lights cannot be automatically dimmed, it will lead to ineffective consumption of energy.

[0004] Similar problems also exist in commercial places. During business hours in shopping malls, supermarkets, etc., the passenger flow distribution in different areas is uneven and changes dynamically with time. For example, some shelf areas are almost unattended during certain periods, but the lighting fixtures still operate at high brightness. In addition, some commercial places keep unnecessary decorative lighting on for a long time to pursue the overall lighting effect, further exacerbating energy waste.

[0005] In public buildings such as schools and hospitals, the lighting in public areas such as corridors and stairwells is often controlled according to a fixed schedule. Even when there are few people late at night, the lights in these areas are often on, which not only consumes a large amount of electrical energy but also increases the loss of the lighting fixtures.

[0006] Based on the above, as people's demands for energy conservation, emission reduction and intelligent life continue to increase, traditional lighting management systems have been difficult to meet the actual needs. Therefore, it is urgent to develop an energy-saving lighting management system that can sense changes in the environment and human activities in real time and accurately control lighting accordingly. So we propose an energy-saving lighting management system based on three-dimensional visualization technology to solve the above existing problems. Summary of the Invention

[0007] 1. Technical Problems to be Solved

[0008] Aiming at the problems existing in the prior art, the purpose of the present invention is to provide an energy-saving lighting management system based on three-dimensional visualization technology. Firstly, the system is based on multi-source data decision-making. Through the construction of a three-dimensional scene construction module, a lighting data acquisition module, and a personnel behavior analysis module, the system can comprehensively collect the spatial information, lighting intensity data, and personnel behavior data of the lighting area. Based on this, the data analysis module can comprehensively consider the function of the lighting area, the current lighting, the probability of personnel activities, and the preset energy-saving target, and formulate a precise optimal lighting strategy, covering the lamp switch state and brightness adjustment level, so as to realize the reasonable allocation of lighting resources. At the same time, the real-time update ability of the three-dimensional model of the three-dimensional scene construction module and the real-time acquisition ability of the data of each module enable the system to adjust the lighting strategy in real time according to the actual changes in the area, ensuring that the lighting always meets the actual needs;

[0009] Then, the introduced simulated annealing algorithm has a powerful energy-saving calculation and optimization ability. It can find the global optimal solution among many possible lighting strategies, rather than being limited to the local optimal solution. This means that the system can minimize energy consumption while meeting the lighting requirements. At the same time, the introduced simulated annealing algorithm can effectively respond to the changes in the usage nature of different time periods and different lighting areas by setting parameters such as the initial temperature, temperature drop rate, and termination temperature, and flexibly adjust the lighting strategy in different scenarios, always maintaining a high-efficiency energy-saving state. In the cost function calculation, the energy consumption and the matching degree of lighting requirements are comprehensively considered. The energy consumption calculation is accurate to the energy consumption of each lamp in different states, and at the same time, the lighting matching degree is calculated by combining the data of the light sensor, and the weight of the two is balanced by adjusting the coefficient to ensure that the lighting quality is not sacrificed while pursuing energy saving.

[0010] 2. Technical solution

[0011] To solve the above problems, the present invention adopts the following technical solutions.

[0012] An energy-saving lighting management system based on three-dimensional visualization technology, including a three-dimensional scene construction module, a lighting data acquisition module, a personnel behavior analysis module, a data analysis module, a lighting control module, and a user interaction module;

[0013] The three-dimensional scene construction module is used to collect the spatial information of the lighting area, including but not limited to building structures, object layouts, personnel activity areas, etc., and construct a high-precision three-dimensional visualization model based on the spatial information. The three-dimensional visualization model has the ability to be updated in real time and can dynamically adjust the display of the three-dimensional model according to the actual changes in the area, such as object movement, personnel entry and exit, etc.

[0014] The light data acquisition module consists of multiple light sensors distributed within the lighting area, which are used to collect real-time light intensity data at different positions and transmit the data to the data analysis module;

[0015] The personnel behavior analysis module collects personnel behavior data through devices such as cameras and infrared sensors installed in the lighting area. The behavior data includes the position, movement trajectory, stay time, etc. of the personnel, and uses artificial intelligence algorithms to analyze the collected behavior data to predict the activity probability of personnel in different areas;

[0016] The data analysis module receives the light intensity data from the light data acquisition module and the personnel behavior data from the personnel behavior analysis module. Based on a preset energy-saving algorithm and combined with a three-dimensional visualization model, it analyzes and obtains the optimal lighting strategy for each lighting area. The energy-saving algorithm comprehensively considers the function of the lighting area, the current light intensity, the personnel activity probability, and the preset energy-saving target. The preset energy-saving target is dynamically adjusted according to different time periods and the usage nature of the lighting area. For example, during non-working hours in the office area, the energy-saving target is set to a higher energy-saving ratio. At the same time, a simulated annealing algorithm is introduced into the energy-saving algorithm to find the global optimal solution among multiple possible lighting strategies and minimize energy consumption to the greatest extent while meeting the lighting requirements. The specific operations are as follows:

[0017] S1. Classification of time and usage nature:

[0018] A day is divided into multiple time periods, such as working hours (8:00 - 18:00), non-working hours (18:00 - 24:00 and 0:00 - 8:00), and the lighting areas are classified into different categories according to their usage nature, such as office areas, corridors, warehouses, etc. Different initial energy-saving target values are set for different category areas and time periods. For example, the energy-saving target for the office area during non-working hours is set to 70%, while the energy-saving target for the corridor during non-working hours is set to 80%;

[0019] S2. Energy-saving optimization process introducing the simulated annealing algorithm:

[0020] S2-1. Initialization: Set the initial temperature T0, the temperature reduction rate α, and the termination temperature T end , and calculate the initial cost function C(S) for each possible lighting strategy S. The cost function comprehensively considers the energy consumption of the lighting area and the degree of meeting the lighting requirements. The energy consumption calculation is as follows;

[0021] S2-2. Suppose there are n lamps in the lighting area, the power of the i-th lamp is P i , and its brightness adjustment coefficient is b i (0 ≤ b i ≤ 1, b i= 1 represents full brightness, b i = 0 represents off), and the on - state of the lighting fixture at a certain moment t is s i (t)(s i (t) = 1 represents on, s i (t) = 0 represents off). Then, within the time period [t1, t2], the energy consumption E(S) calculation formula for this lighting area is:

[0022]

[0023] where Δt is the time interval;

[0024] S2 - 3. The matching degree L(S) between the average light level of the lighting area and the preset light level is calculated through the data of the light sensors. Suppose there are m light sensors in this area, and the actual light intensity measured by the j - th sensor is I j , and the preset light intensity is I 0j . Then the calculation formula for the matching degree L(S) is:

[0025]

[0026] Combining the energy consumption and the light matching degree, the cost function C(S) is:

[0027] C(S)=E(S)+k(1 - L(S))

[0028] where k is an adjustment coefficient used to balance the weights of energy consumption and light requirements;

[0029] S2 - 4. Iteration process: At the current temperature T, perform a neighborhood operation on the current lighting strategy S to generate a new lighting strategy S′, and calculate the cost function C(S′) of the new strategy. According to the Metropolis criterion, that is, if ΔC = C(S′)-C(S)<0, then accept the new strategy; otherwise, accept the new strategy with a probability of e -ΔC / T The neighborhood operation can be to slightly adjust the on - off state or brightness level of some lighting fixtures. For example, randomly select a lighting fixture and change its on - off state or adjust its brightness level within a certain range;

[0030] S2 - 5. Temperature update: Update the temperature according to the formula T = T×α until the temperature is lower than T end ;

[0031] S2 - 6. Final strategy determination: The finally obtained lighting strategy is the optimal strategy at the end of the iteration of the simulated annealing algorithm. This strategy minimizes the energy consumption to the greatest extent while meeting the lighting requirements;

[0032] S3. Data update and feedback adjustment:

[0033] During the operation of the system, according to the actual energy consumption data and personnel behavior data, the energy-saving target values for different time periods and usage natures are updated regularly. For example, if it is found that the personnel activities in a certain area during a certain period are frequently beyond expectations, the energy-saving target for that period can be appropriately reduced, and vice versa. At the same time, according to the updated energy-saving target, the simulated annealing algorithm is re-run to optimize and adjust the current lighting strategy. The optimal lighting strategy includes the on / off state of the lamps, brightness adjustment levels, etc.;

[0034] The lighting control module, based on the optimal lighting strategy obtained by the data analysis module, uses wireless communication technologies such as ZigBee and Wi-Fi to control the intelligent lamps distributed in the lighting area in real time, realizing precise switching and brightness adjustment of the lamps to achieve the purpose of energy saving;

[0035] The user interaction module provides a graphical user interface. Through the graphical user interface, the user can view the 3D visualization model, real-time lighting data, personnel behavior information, and the current lighting strategy. At the same time, the user can manually adjust the lighting strategy according to actual needs. The system will record the user's adjustment operations and feedback them to the data analysis module for optimizing the subsequent generation of lighting strategies.

[0036] Furthermore, the 3D scene construction module uses a combination of laser scanning technology and computer vision technology to collect spatial information. Among them, laser scanning technology is used to obtain the accurate 3D coordinate information of building structures and large objects, and computer vision technology processes the images collected by multiple cameras to identify details such as the texture and material of objects, constructing a more realistic and accurate 3D visualization model.

[0037] Furthermore, the light data collection module also includes an automatic calibration sub-module. The automatic calibration sub-module regularly calibrates the light sensors distributed in the lighting area using a standard light source. The calibration process uses a combination of spectral matching technology and the least squares method. The spectral matching technology is used to compare the spectral characteristics of the standard light source and the sensor measurement, and the least squares method is used to optimize the calibration parameters, improving the accuracy and stability of the light intensity data collection, ensuring the reliability of the data obtained by the data analysis module, and thus enhancing the scientific nature of lighting strategy formulation.

[0038] Furthermore, the artificial intelligence algorithm of the personnel behavior analysis module uses a recurrent neural network (RNN) in deep learning, such as a long short-term memory network (LSTM), to learn and analyze the historical behavior data of personnel, improving the accuracy of predicting the probability of personnel activities. The training data includes not only the personnel behavior data collected in this lighting area but also integrates the personnel behavior data of similar types of lighting areas to enhance the generalization ability of the model.

[0039] Furthermore, the wireless communication technology of the lighting control module adopts an adaptive channel selection technology, which real-time monitors the signal strength, interference situation, and data transmission rate of ZigBee and Wi-Fi communication channels, and dynamically selects the optimal channel by using a reinforcement learning algorithm. The reinforcement learning algorithm continuously interacts with the communication environment, and according to the channel state feedback reward value, gradually learns the best channel selection strategy in different scenarios, effectively avoiding communication interference, ensuring the stable transmission of lighting control instructions, and realizing the precise control of intelligent lamps.

[0040] Furthermore, the user interaction module also captures the eye movement trajectory and gesture actions of the user during the operation of the graphical user interface through eye movement tracking technology and gesture recognition technology, analyzes these data by using machine learning algorithms, judges the operation intention and focus of the user, and automatically adjusts the interface layout and information display method. For example, if the user frequently pays attention to the lighting strategy adjustment in a certain area, the relevant information of this area will be placed in a more prominent position on the interface, improving the convenience of user operation and the interaction experience. At the same time, it provides user behavior preference data for the data analysis module to optimize the generation of lighting strategies.

[0041] Furthermore, the data analysis module uses blockchain technology to store and analyze energy consumption data. Each energy consumption data record is accompanied by a timestamp and a digital signature, is packaged into a block and linked into a chain. By using the immutable and traceable characteristics of the blockchain, the authenticity and integrity of the energy consumption data are ensured. At the same time, a federated learning algorithm is used to collaboratively analyze the energy consumption data on multiple distributed nodes. Each node jointly trains the model without sharing the original data, mines the energy consumption patterns and energy-saving potential in different regions and different time periods, and provides comprehensive data support for formulating more precise energy-saving goals and lighting strategies.

[0042] 3. Beneficial Effects

[0043] Compared with the prior art, the advantages of the present invention are as follows:

[0044] (1) In this solution, the system is based on multi-source data decision-making. By building a three-dimensional scene construction module, a light data acquisition module, and a personnel behavior analysis module, the system can comprehensively collect the spatial information, light intensity data, and personnel behavior data of the lighting area. Based on this, the data analysis module can comprehensively consider the function of the lighting area, the current light, the probability of personnel activities, and the preset energy-saving goal, and formulate a precise optimal lighting strategy, covering the lamp switch state and brightness adjustment level, realizing the reasonable allocation of lighting resources. At the same time, the real-time update ability of the three-dimensional model of the three-dimensional scene construction module and the real-time acquisition ability of the data of each module enable the system to adjust the lighting strategy in real time according to the actual changes in the area, ensuring that the lighting always meets the actual needs;

[0045] (2) In this solution, the introduced simulated annealing algorithm has powerful energy-saving calculation and optimization capabilities. It can find the global optimal solution among numerous possible lighting strategies instead of being limited to local optima. This means that the system can minimize energy consumption while meeting lighting requirements. At the same time, the introduced simulated annealing algorithm sets parameters such as the initial temperature, temperature drop rate, and termination temperature, and accepts new strategies according to the Metropolis criterion during the iteration process. It can effectively cope with changes in usage nature in different time periods and different lighting areas, flexibly adjust lighting strategies in different scenarios, and always maintain a highly energy-saving state. In the cost function calculation, both energy consumption and lighting demand matching degree are comprehensively considered. The energy consumption calculation is accurate to the energy consumption of each lamp in different states. At the same time, the lighting matching degree is calculated in combination with the data of the light sensor, and the weights of the two are balanced through adjustment coefficients to ensure that lighting quality is not sacrificed while pursuing energy conservation. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a schematic diagram of the system module composition and data flow between modules of the present invention;

[0047] Figure 2 It is a schematic diagram briefly describing the principle of the three-dimensional scene construction module of the present invention;

[0048] Figure 3 It is a schematic diagram briefly describing the principle of the light data acquisition module of the present invention;

[0049] Figure 4 It is a schematic diagram briefly describing the principle of the personnel behavior analysis module of the present invention;

[0050] Figure 5 It is a schematic diagram briefly describing the principle of the data analysis module of the present invention;

[0051] Figure 6 It is a schematic diagram briefly describing the principle of the lighting control module of the present invention;

[0052] Figure 7 It is a schematic diagram briefly describing the principle of the user interaction module of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0053] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings of the present invention; obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0054] Embodiment 1:

[0055] This embodiment takes an office space as an example and combines the accompanying drawings of the specification Figures 1-7Schematic diagram, which details the operation process of the energy-saving lighting management system based on 3D visualization technology.

[0056] I. 3D Scene Construction:

[0057] Use laser scanning technology to obtain accurate 3D coordinate information of building structures and large objects, and use computer vision technology to process images collected by multiple cameras to identify details such as the texture and material of objects, and construct a realistic and accurate 3D visualization model. For example, in the office area, accurately scan the positions of large objects such as desks and filing cabinets, determine the wall color and material through image recognition, and construct a 3D scene that can be updated in real time. When people enter or leave or objects move, the model is dynamically adjusted and displayed accordingly.

[0058] II. Lighting Data Collection:

[0059] Deploy lighting sensors in the lighting area, and the automatic calibration sub-module regularly calibrates using a standard light source. The calibration process combines spectral matching technology and the least squares method to improve the accuracy and stability of data collection. Suppose 5 lighting sensors are set in the office area, and the spectral characteristics of the standard light source are known. Compare the measured spectrum with the standard spectrum through spectral matching technology, and use the least squares method to optimize the calibration parameters. The collected lighting intensity data is transmitted to the data analysis module in real time.

[0060] III. Personnel Behavior Analysis:

[0061] Install cameras and infrared sensors in the lighting area to collect personnel behavior data, and use long short-term memory network (LSTM) to learn and analyze historical behavior data to predict the probability of personnel activities. For example, during office hours, count data such as the number of people entering and leaving and the staying time in a certain office area, and train the model in combination with the personnel behavior data of similar office areas. After training, predict the probability of personnel activities in this area when there is an emergency overtime project during non-office hours.

[0062] IV. Data Analysis and Optimal Lighting Strategy Formulation:

[0063] 1. Classification by Time and Usage Nature: Divide a day into office hours (8:00 - 18:00) and non-office hours (18:00 - 24:00 and 0:00 - 8:00). The energy-saving target for the office area during non-office hours is set at 70%, and the energy-saving target for the corridor during non-office hours is set at 80%.

[0064] Energy-saving optimization process introducing the simulated annealing algorithm:

[0065] 2. Initialization: Set the initial temperature T0 = 100, the temperature reduction rate α = 0.95, and the termination temperature T end= 1. Assume there are 10 lamps in the lighting area, and the power P of the i-th lamp i is 20W, 25W, 30W, …, 45W respectively, and the initial brightness adjustment factor b i is 1 (full brightness) for all, and the lamp on-state s i (t) is 1 (on) initially for all. Calculate the initial cost function C(S), where the energy consumption E(S) and the light matching degree L(S) are calculated according to the following formulas, and assume the adjustment factor k = 5.

[0066] 3. Energy consumption calculation: Assume the time interval Δt = 1h. During the time period [t1, t2] (such as 17:00 - 18:00), according to the formula calculate the energy consumption of the lighting area. For example, during this time period, if the states of all lamps remain unchanged, then E(S) = (20×1×1 + 25×1×1 + … + 45×1×1)×1 = 325Wh.

[0067] 4. Light matching degree calculation: Assume there are 3 light sensors in the office area, and the actual light intensity J measured by the j-th sensor j is 300lux, 350lux, 400lux respectively, and the preset light intensity O j is 350lux for all. According to the formula calculate the light matching degree

[0068] 5. Cost function calculation: Considering the comprehensive energy consumption and light matching degree, the cost function C(S) = E(S) + k(1 - L(S)) = 325 + 5×(1 - 0.93) = 325.35.

[0069] 6. Iteration process: At the current temperature T = T0, perform a neighborhood operation on the current lighting strategy S. For example, randomly select the 3rd lamp and adjust its brightness adjustment factor from 1 to 0.8 to generate a new lighting strategy S'. Recalculate the cost function C(S′) of the new strategy. Assume that the energy consumption under the new strategy becomes E(S′) = 300Wh and the light matching degree becomes L(S′) = 0.9. Then C(S′) = 300 + 5×(1 - 0.9) = 300.5. Since C(S′) - C(S) = 300.5 - 325.35 = -24.85 < 0, according to the Metropolis criterion, accept the new strategy.

[0070] 7. Temperature update: Update the temperature according to the formula T new = αT old That is, T new = 0.95×100 = 95. Repeat the iteration process until the temperature is lower than T end .

[0071] 8. Final strategy determination: The optimal strategy at the end of the iteration of the simulated annealing algorithm is the final lighting strategy, which minimizes energy consumption to the greatest extent while meeting the lighting requirements.

[0072] 9. Data update and feedback adjustment: During the operation of the system, if it is found that the personnel activities in the office area are frequently beyond expectations during a certain period of non-office hours, for example, if it was originally predicted that only 2 people would work overtime during this period, but actually 8 people work overtime, then the energy-saving target for this period is appropriately reduced, such as from 70% to 60%, and the simulated annealing algorithm is re-run to optimize the current lighting strategy.

[0073] V. Lighting control:

[0074] The lighting control module, according to the optimal lighting strategy obtained by the data analysis module, adopts adaptive channel selection technology to continuously monitor the signal strength, interference situation, and data transmission rate of ZigBee and Wi-Fi communication channels, and dynamically selects the optimal channel using a reinforcement learning algorithm, and performs real-time control of intelligent lamps through wireless communication technology. For example, when it detects that the signal strength of the ZigBee channel weakens and the interference increases, it switches to the Wi-Fi channel to achieve precise switching and brightness adjustment of the lamps, thus achieving the purpose of energy saving.

[0075] VI. User interaction:

[0076] Users can view the 3D visualization model, real-time lighting data, personnel behavior information, and the current lighting strategy through the graphical user interface. If users frequently pay attention to the lighting strategy adjustment in a certain area of the office area, the user interaction module captures user operations through eye movement tracking technology and gesture recognition technology (this is an existing publicly known technology, so it will not be described in detail here), analyzes using machine learning algorithms, places the relevant information of this area in a more prominent position on the interface. At the same time, the operations of users manually adjusting the lighting strategy are recorded and fed back to the data analysis module for optimizing the generation of subsequent lighting strategies.

[0077] Embodiment 2:

[0078] This embodiment takes a large shopping mall as an example and, in combination with the schematic diagrams attached to the specification Figures 1-7 details the operation process of this energy-saving lighting management system based on 3D visualization technology.

[0079] I. 3D scene construction:

[0080] Use laser scanning technology to accurately measure the building structure of the shopping mall, such as the positions and shapes of walls and columns, as well as the layouts of large fixed facilities, such as elevators and escalators, to obtain their three-dimensional coordinate information. At the same time, process the images collected by cameras distributed in various corners of the shopping mall through computer vision technology to identify the textures and material details of objects such as product shelves and display stands. For example, in the clothing area, construct a realistic three-dimensional scene by identifying the colors, materials, and product display methods of the shelves. When the shelf position is adjusted or the products are changed in display, the three-dimensional scene construction module can perceive and update the three-dimensional visualization model in real time to ensure that the model is consistent with the actual scene of the shopping mall.

[0081] II. Illumination data collection:

[0082] Install illumination sensors in different areas of the shopping mall, such as setting 5 in the clothing area, 3 in the dining area, and 4 in the electrical appliance area. The automatic calibration sub-module in the illumination data collection module regularly (such as at 3 am every day) calibrates these sensors using a standard light source. During the calibration process, spectral matching technology compares the spectral characteristics of the standard light source with those measured by the sensors. Assuming the spectral distribution of the standard light source is known and there is a certain deviation between the spectrum measured by the sensor and the standard spectrum, optimize the calibration parameters through the least squares method. After calibration, the illumination sensors collect illumination intensity data in real time and transmit it to the data analysis module. For example, at a certain moment, the illumination intensities measured by 5 sensors in the clothing area are 350 lux, 400 lux, 380 lux, 420 lux, and 360 lux respectively.

[0083] III. Personnel behavior analysis:

[0084] Collect personnel behavior data through cameras and infrared sensors installed throughout the shopping mall. Use long short-term memory network (LSTM) to learn and analyze historical personnel behavior data to predict the probability of personnel activities. The training data includes not only the personnel behavior data in different areas of this shopping mall but also integrates the personnel behavior data of other shopping malls with similar scale and layout to enhance the generalization ability of the model. For example, from 2 pm to 4 pm on weekends, count data such as the number of people entering and leaving and the staying time in the dining area, and train in combination with the data of the same period in previous weekends and the data of the dining areas of other shopping malls on weekends. After training, predict the distribution probability of crowded and sparse areas in the dining area at the same time period next week.

[0085] IV. Data analysis and formulation of the best lighting strategy:

[0086] 1. Classification by time and usage nature: The business hours of the shopping mall are divided into different time periods, such as peak hours (11:00 - 14:00, 17:00 - 20:00) and off-peak hours (the remaining business hours). According to the usage nature of each area in the shopping mall, the energy-saving target for the clothing area and the electrical appliance area is set at 60% during off-peak hours, the energy-saving target for the dining area is set at 70% during non-peak dining hours (14:00 - 17:00), and the energy-saving target for the corridor is set at 80% during off-peak hours.

[0087] Energy-saving optimization process introducing the simulated annealing algorithm:

[0088] 2. Initialization: Set the initial temperature T0 = 120, the temperature reduction rate α = 0.96, and the termination temperature T end = 2. Assume there are 8 lamps in the clothing area, and the power P i of the i-th lamp are 30W, 35W, 40W, 32W, 38W, 42W, 36W, 45W respectively. The initial brightness adjustment coefficient b i are all 1 (full brightness), and the lamp on-off state s i (t) are initially all 1 (on). Let the adjustment coefficient k = 6, and calculate the initial cost function C(S), where the energy consumption E(S) and the lighting matching degree L(S) are calculated according to the following formulas.

[0089] 3. Energy consumption calculation: Let the time interval Δt = 1h. During the time period [t1, t2] (such as 15:00 - 16:00), according to the formula calculate the energy consumption of the lighting area. During this time period, the states of each lamp remain unchanged, so E(S) = (30×1×1 + 35×1×1 + … + 45×1×1)×1 = 308Wh.

[0090] 4. Lighting matching degree calculation: Assume the actual light intensities J j measured by 5 light sensors in the clothing area are 380lux, 420lux, 360lux, 400lux, 390lux respectively, and the preset light intensity O j are all 400lux. According to the formula calculate the lighting matching degree

[0091] 5. Cost function calculation: Combining energy consumption and lighting matching degree, the cost function C(S) = E(S) + k(1 - L(S)) = 308 + 6×(1 - 0.95) = 308.3.

[0092] 6. Iterative process: At the current temperature T = T0, perform a neighborhood operation on the current lighting strategy S. For example, randomly select the 5th lamp and adjust its brightness adjustment coefficient from 1 to 0.7 to generate a new lighting strategy S'. Recalculate the cost function C(S′) of the new strategy. Assume that the energy consumption under the new strategy becomes E(S′) = 280 Wh and the lighting matching degree becomes L(S′) = 0.92. Then C(S′) = 280 + 6×(1 - 0.92) = 280.48. Since C(S′) - C(S) = 280.48 - 308.3 = -27.82 < 0, according to the Metropolis criterion, accept the new strategy.

[0093] 7. Temperature update: Update the temperature according to the formula T new = αT old That is, T new = 0.96×120 = 115.2. Repeat the iterative process until the temperature is lower than T end .

[0094] 8. Final decision determination: The optimal strategy at the end of the simulated annealing algorithm iteration is the final lighting strategy, which minimizes the energy consumption to the greatest extent while meeting the lighting requirements.

[0095] 9. Data update and feedback adjustment: During the system operation, if it is found that the personnel activities in the electrical appliance area during a certain non-peak period are more frequent than expected. For example, it was originally predicted that there would be about 10 people browsing during this period, but actually there are 30 people. The energy-saving target for this period can be appropriately reduced, such as from 60% to 50%, and the simulated annealing algorithm can be re-run to optimize and adjust the current lighting strategy.

[0096] V. Lighting control:

[0097] The lighting control module adopts adaptive channel selection technology according to the best lighting strategy obtained by the data analysis module. It real-time monitors the signal strength, interference situation, and data transmission rate of ZigBee and Wi-Fi communication channels, and dynamically selects the optimal channel using the reinforcement learning algorithm. For example, in a certain area of a shopping mall, the Wi-Fi channel is severely interfered by surrounding devices, the signal strength weakens, and the data transmission rate decreases. The reinforcement learning algorithm feeds back the reward value according to the channel state and selects the ZigBee channel as the communication channel. Through wireless communication technology, it realizes real-time control of intelligent lamps, achieving precise switching and brightness adjustment of lamps and achieving the purpose of energy saving.

[0098] VI. User interaction:

[0099] The mall management staff can view the 3D visualization model, real-time lighting data, personnel behavior information, and the current lighting strategy through a graphical user interface. If the management staff frequently pays attention to the adjustment of the lighting strategy in the dining area, the user interaction module captures the operations through eye-tracking technology and gesture recognition technology (this is an existing known public technology, so no detailed description will be given here), analyzes using machine learning algorithms, and places the information related to the dining area in a more prominent position on the interface. At the same time, the operations of the management staff manually adjusting the lighting strategy are recorded and fed back to the data analysis module for optimizing the generation of subsequent lighting strategies.

[0100] The above is only a preferred specific embodiment of the present invention; however, the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution of the present invention and its improved concept, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

Claims

1. An energy-saving lighting management system based on 3D visualization technology, comprising a 3D scene construction module, a lighting data acquisition module, a personnel behavior analysis module, a data analysis module, a lighting control module, and a user interaction module, characterized in that: The 3D scene construction module is used to collect spatial information of the lighting area, including but not limited to building structures, object layouts, personnel activity areas, etc., and construct a high-precision 3D visualization model based on the spatial information. The 3D visualization model has the ability to be updated in real time and can dynamically adjust the display of the 3D model according to actual changes in the area, such as object movement, personnel entry and exit, etc. The lighting data acquisition module is composed of multiple lighting sensors distributed in the lighting area, which is used to collect the lighting intensity data at different positions in real time and transmit the data to the data analysis module. The personnel behavior analysis module collects the behavior data of personnel through devices such as cameras and infrared sensors set in the lighting area. The behavior data includes the position, movement trajectory, stay time, etc. of the personnel, and uses artificial intelligence algorithms to analyze the collected behavior data to predict the activity probability of personnel in different areas. The data analysis module receives the lighting intensity data from the lighting data acquisition module and the personnel behavior data from the personnel behavior analysis module. Based on the preset energy-saving algorithm, combined with the 3D visualization model, it analyzes and obtains the best lighting strategy for each lighting area. The energy-saving algorithm comprehensively considers the function of the lighting area, the current lighting intensity, the personnel activity probability, and the preset energy-saving target. The preset energy-saving target is dynamically adjusted according to different time periods and the usage nature of the lighting area. For example, during non-working hours in the office area, the energy-saving target is set to a higher energy-saving ratio. At the same time, a simulated annealing algorithm is introduced into the energy-saving algorithm to find the global optimal solution among multiple possible lighting strategies, and minimize energy consumption to the greatest extent under the premise of meeting lighting requirements. The specific operations are as follows: S1. Classification of time and usage nature: A day is divided into multiple time periods, such as office hours (8:00 - 18:00), non-office hours (18:00 - 24:00 and 0:00 - 8:00), and the lighting areas are divided into different categories according to their usage nature, such as office areas, corridors, warehouses, etc. Different initial energy-saving target values are set for different category areas and time periods. For example, the energy-saving target for the office area during non-office hours is set to 70%, while the energy-saving target for the corridor during non-office hours is set to 80%. S2. Energy-saving optimization process introducing the simulated annealing algorithm: S2-1. Initialization: Set the initial temperature T0, the temperature drop rate α, and the termination temperature T end , and calculate the initial cost function C(S) for each possible lighting strategy S. The cost function comprehensively considers the energy consumption of the lighting area and the degree of meeting the lighting requirements. The energy consumption calculation is as follows; S2-2. Suppose there are n lamps in the lighting area, and the power of the i-th lamp is P i , and its brightness adjustment coefficient is b i (0 ≤ b i ≤ 1, b i = 1 indicates full brightness, b i = 0 indicates off), and the on-state of the lamp at a certain moment t is s i (t) (s i (t) = 1 indicates on, s i (t) = 0 indicates off). Then, within the time period [t1, t2], the calculation formula for the energy consumption E(S) of this lighting area is: where Δt is the time interval; S2-3. The matching degree L(S) between the average illumination level of the illumination area and the preset illumination level is calculated through the data of the light sensors. Suppose there are m light sensors in this area, and the actual illumination intensity measured by the j-th sensor is I j , and the preset illumination intensity is I 0j . Then the calculation formula for the matching degree L(S) is as follows: Combining the comprehensive energy consumption and lighting matching degree, the cost function C(S) is: C(S) = E(S) + k(1 - L(S)) where k is an adjustment coefficient used to balance the weights of energy consumption and lighting requirements; S2-4. Iterative process: At the current temperature T, perform a neighborhood operation on the current lighting strategy S to generate a new lighting strategy S′, and calculate the cost function C(S′) of the new strategy. According to the Metropolis criterion, that is, if ΔC = C(S′) - C(S) < 0, then accept the new strategy; otherwise, with probability e -ΔC / T Accept the new strategy. The neighborhood operation can be to slightly adjust the on / off state or brightness level of some lamps. For example, randomly select a lamp and change its on / off state or adjust its brightness level within a certain range; S2-5, Temperature Update: Update the temperature according to the formula T = T × α until the temperature is lower than T end ; S2 - 6. Determination of the final strategy: The finally obtained lighting strategy is the optimal strategy when the simulated annealing algorithm iteration ends. This strategy minimizes energy consumption to the greatest extent under the premise of meeting lighting requirements. S3. Data update and feedback adjustment: During the operation of the system, according to the actual energy consumption data and personnel behavior data, the energy-saving target values for different time periods and usage natures are updated regularly. For example, if it is found that the personnel activities in a certain area during a certain period are frequently beyond expectations, the energy-saving target for that period can be appropriately reduced; otherwise, the energy-saving target can be increased. At the same time, according to the updated energy-saving target, the simulated annealing algorithm is re-run to optimize and adjust the current lighting strategy. The optimal lighting strategy includes the on / off state of the lamps, the brightness adjustment level, etc.; The lighting control module, according to the optimal lighting strategy obtained by the data analysis module, uses wireless communication technologies such as ZigBee and Wi-Fi to perform real-time control on the intelligent lamps distributed in the lighting area, realizing the precise on / off and brightness adjustment of the lamps to achieve the purpose of energy saving; The user interaction module provides a graphical user interface. Through the graphical user interface, the user can view the 3D visualization model, real-time lighting data, personnel behavior information, and the current lighting strategy. At the same time, the user can manually adjust the lighting strategy according to actual needs. The system will record the user's adjustment operations and feedback them to the data analysis module for optimizing the subsequent generation of lighting strategies.

2. The energy-saving lighting management system based on three-dimensional visualization technology according to claim 1, wherein The 3D scene construction module uses a combination of laser scanning technology and computer vision technology to collect spatial information. Among them, laser scanning technology is used to obtain the precise 3D coordinate information of the building structure and large objects. Computer vision technology processes the images collected by multiple cameras to identify details such as the texture and material of the objects, and constructs a more realistic and accurate 3D visualization model.

3. The energy-saving lighting management system based on three-dimensional visualization technology according to claim 1, characterized in that The light data collection module also includes an automatic calibration sub-module. The automatic calibration sub-module regularly calibrates the light sensors distributed in the lighting area using a standard light source. The calibration process uses a combination of spectral matching technology and the least squares method. The spectral matching technology is used to compare the spectral characteristics of the standard light source and the sensor measurement. The least squares method is used to optimize the calibration parameters, improve the accuracy and stability of the light intensity data collection, ensure the reliability of the data obtained by the data analysis module, and thus enhance the scientificity of the lighting strategy formulation.

4. An energy-saving lighting management system based on 3D visualization technology according to claim 1, characterized in that: The artificial intelligence algorithm of the personnel behavior analysis module uses a recurrent neural network (RNN) in deep learning, such as a long short-term memory network (LSTM), to learn and analyze the historical behavior data of personnel, improving the accuracy of predicting the probability of personnel activities. The training data includes not only the personnel behavior data collected in this lighting area but also integrates the personnel behavior data of similar types of lighting areas to enhance the generalization ability of the model.

5. An energy-saving lighting management system based on 3D visualization technology according to claim 1, characterized in that: The wireless communication technology of the lighting control module adopts adaptive channel selection technology, which monitors the signal strength, interference situation and data transmission rate of ZigBee and Wi-Fi communication channels in real time, and dynamically selects the optimal channel by using the reinforcement learning algorithm. The reinforcement learning algorithm continuously interacts with the communication environment, obtains the reward value according to the channel state feedback, and gradually learns the best channel selection strategy in different scenarios, effectively avoiding communication interference, ensuring the stable transmission of lighting control instructions, and realizing the precise control of intelligent lamps.

6. The energy-saving lighting management system based on 3D visualization technology according to claim 1, wherein: The user interaction module also captures the eye movement trajectory and gesture actions of the user during the operation of the graphical user interface through eye movement tracking technology and gesture recognition technology, analyzes this data by using machine learning algorithms, judges the operation intention and focus of the user, and automatically adjusts the interface layout and information display method. For example, if the user frequently pays attention to the adjustment of the lighting strategy in a certain area, the relevant information of this area will be placed in a more prominent position on the interface, improving the convenience of user operation and the interaction experience. At the same time, it provides user behavior preference data for the data analysis module to optimize the generation of lighting strategies.

7. An energy-saving lighting management system based on 3D visualization technology according to claim 1, characterized in that: The data analysis module uses blockchain technology to store and analyze energy consumption data. Each energy consumption data record is accompanied by a timestamp and a digital signature, is packaged into a block and linked into a chain. By using the immutable and traceable characteristics of the blockchain, the authenticity and integrity of the energy consumption data are ensured. At the same time, the federated learning algorithm is used to collaboratively analyze the energy consumption data on multiple distributed nodes. Each node jointly trains the model without sharing the original data, mines the energy consumption patterns and energy-saving potentials in different regions and different time periods, and provides comprehensive data support for formulating more precise energy-saving goals and lighting strategies.

Images