A method for optimizing the layout of photovoltaic sensors based on the differential evolution algorithm

Abstract

The present invention claims protection for an optimization layout method of photovoltaic sensors based on the differential evolution algorithm, which includes determining a model for the average illuminance level and illuminance uniformity (Uo) of lighting lamps; taking the energy consumption of lighting as the objective function and the illuminance level and uniformity as the constraint conditions for optimization design; using the differential evolution optimization algorithm (DE) to solve the problem of the layout of the optical sensors; determining the optimal number and the optimal positions of the sensors based on the average dimming level and the optimal illuminance value region solved by the differential evolution algorithm. It aims to minimize costs (i.e., sensors and electric energy), improve visual comfort and controller performance (i.e., reduce complexity). This method has superior performance in terms of small computational amount and optimal solutions (i.e., the number and positions of the optical sensors). In addition, this method has strong practicability for the implementation of wired and wireless optical sensors in small and large buildings, new buildings and renovation projects.

Description

Technical Field

[0001] The present invention belongs to the technical field of building energy, and particularly relates to an optimized layout method of photovoltaic sensors based on a differential evolution algorithm. Background Technique

[0002] The energy used in buildings globally accounts for a large portion of the total energy, approximately 40% of the total energy and about 30% of the carbon dioxide emissions, thus leading to greenhouse gas effects such as climate change and global warming. Lighting is one of the main subsystems of the building energy system, accounting for about 20%-45% of the total energy consumption.

[0003] The lighting system can achieve higher energy-saving potential in buildings by implementing several strategies, such as using energy-saving lamp technology, designing appropriate illuminance levels, and dimming control. For dimming control, the light-sensing-based control strategy is the main strategy for increasing building lighting energy conservation. Therefore, the layout of light sensors in terms of quantity and location is crucial for the cost of the sensors and the performance of the control system. Improper placement of light sensors results in poor energy performance of the system and fails to achieve an energy-saving building. The layout methods of light sensors can be divided into three types: fixed (i.e., sensors are juxtaposed on the lamps), mathematical functions, and optimization-based methods.

[0004] Most studies consider the layout problem of light sensors based on mathematical functions. These methods result in a larger number of sensors and the positions of the sensors are not optimal. Therefore, it increases the initial cost of the light sensors and the lighting system operates under less than ideal conditions. On the other hand, once the number of sensors exceeds the number of lamps, conflicts occur in the central controller, and the sensors provide illuminance information to the controller as inputs. In fact, in the lighting system design, the number of sensors (N s ) refers to the number of logical areas equal to or less than the number of lamps (N l ), and can be mathematically expressed as N s ≤N l . In addition, for the layout of light sensors, no optimal method has been observed by taking illuminance uniformity as a design parameter.

[0005] Differential evolution algorithm (DE) is a well-known population-based stochastic optimization technique, which has advantages such as few parameter settings, short calculation time, and free derivative algorithms. DE has been widely applied to heating, ventilation, and air conditioning (HVAC) systems. For example, optimizing the temperature set points of areas and chilled water in a room. In fact, in lighting systems, DE has been used to optimize the dimming levels of artificial lighting and luminaire layout design. Therefore, this study selects the differential evolution algorithm as the optimization technique for the dimming level of LED luminaires in the light sensor layout problem. Most previous work has focused on the use of traditional optimization methods, such as convex optimization, linear programming, and iterative methods. However, traditional optimization methods have significant drawbacks when dealing with complex problems and large systems, such as a large amount of computation.

[0006] Therefore, a new method for optimizing the layout of light sensors is designed. Using the differential evolution optimization algorithm, the optimization of the lighting system control strategy is achieved by minimizing the dimming level of luminaires, thereby minimizing lighting energy consumption, maintaining the visual comfort of the building, and reducing the initial cost of sensors at the same time.

[0007] After retrieval, the application publication number CN104156584A, a sensor target allocation method of a multi-objective optimization differential evolution algorithm, the method includes: calculating the importance degree of the target according to the target information, establishing a sensor target allocation constraint multi-objective optimization function, encoding the allocation scheme and generating the initial population chromosome, generating the offspring scheme population using the differential evolution algorithm, merging and screening the population, and obtaining the Pareto front solution set of the allocation scheme, etc. The present invention combines the characteristics of the differential evolution algorithm, which is simple and easy to use, has good robustness, and has strong global search ability in group difference heuristic random search, and provides a Pareto set multi-objective optimization allocation strategy; on the basis of the sensor target monitoring efficiency function, the sensor utilization rate function is added, and the allocation problem is transformed into a multi-objective optimization problem, which can save sensor resources as much as possible under the condition of meeting the monitoring accuracy requirements and realize the reasonable and effective allocation of sensor resources. This technology only starts from the perspective of solving the optimization function. Although it optimizes the allocation of sensors, it increases the number of objective functions, increasing the difficulty and time of solving the problem.

[0008] Before establishing the optimization function, in the control strategy, pulse width modulation (PWM) is used to control the dimming level, minimizing the energy consumption in advance and reducing the difficulty in solving the optimization problem.

[0009] CN110062389A, a method for optimizing the deployment of sensor network nodes based on an improved differential evolution algorithm, which has been successfully applied to the node optimization deployment of wireless sensor networks. By setting chaotic mapping population initialization, the diversity of the initial population is improved; the elite population is used to guide the mutation vector, accelerating the global optimization speed of the population; and a parameter adaptive adjustment mechanism is used to enhance the adaptability of the algorithm to nodes. The advantages of the present invention are as follows: compared with the basic differential evolution algorithm, the improved algorithm has a greater degree of improvement in node coverage rate and convergence speed, and can effectively handle possible emergencies of nodes, enhancing the adaptability of the algorithm; the improved differential evolution algorithm effectively avoids the population falling into local optimum and improves the optimization ability of the algorithm. Compared with the differential evolution algorithm before improvement, the network coverage rate is increased by about 5%, meeting the coverage requirements of the monitoring area, accelerating the convergence speed, and the improved algorithm has strong adaptability. This technology did not consider that the sensing range of the network in actual sensing is in an irregular state, and this state will affect the coverage quality of routing nodes. However, the present invention considers the nodes in actual sensing, and the coverage range is the actual range. Summary of the Invention

[0010] The present invention aims to solve the above problems of the prior art. A method for optimizing the layout of photovoltaic sensors based on a differential evolution algorithm is proposed. The technical solution of the present invention is as follows:

[0011] A method for optimizing the layout of photovoltaic sensors based on a differential evolution algorithm, which includes the following steps:

[0012] Determine the model of the average illuminance level and illuminance uniformity of the lighting lamps; use the energy consumption of lighting as the objective function, and use the illuminance level and uniformity as the constraint conditions for optimization design; use the differential evolution optimization algorithm to solve the problem of the layout of the optical sensors. The improvement of the differential evolution optimization algorithm applied to solving the problem of the layout of the optical sensors lies in its strong robustness in the optimization problem of the non-linear function based on the layout of the optical sensors. Under the same accuracy requirements, the differential evolution algorithm has a fast convergence speed; determine the optimal number of sensors and the optimal positions of the sensors based on the average dimming level and the optimal illuminance value area obtained by solving based on the differential evolution algorithm.

[0013] Further, the average illuminance level of the lighting lamps and the illuminance uniformity U o The model of is:

[0014]

[0015] where Nl is the number of lamps, n l is the number of lamps in each lamp, Φ nFor the output of the lamp, UF is the utilization factor, MF is the maintenance factor, BF is the ballast factor, and A is the surface area of the room;

[0016]

[0017] where E min is the minimum illuminance value of the room.

[0018] Furthermore, in order to determine the value of UF, it is necessary to first calculate the room index RI, and the formula for RI is expressed as follows:

[0019]

[0020] where L and W are the length and width of the room respectively, and H is the vertical distance from the lamp to the working plane; after obtaining the RI value, the UF value can be determined by referring to the lamp data sheet provided by the lamp manufacturer.

[0021] Furthermore, when the lamp uses an LED lamp, in terms of the control strategy, pulse width modulation PWM is used to control the dimming level. Under PWM control, it is assumed that the dimming level of the LED lamp is linearly related to its output power; reducing the dimming degree of the lamp to the lowest means reducing the electric energy to the lowest; in order to find the optimal arrangement of the light sensor corresponding to the minimum energy consumption, a problem formula based on minimizing the dimming level of the LED lamp is established.

[0022] Furthermore, taking the energy consumption of lighting as the objective function and the illuminance level and uniformity as the constraint conditions for optimization design, specifically including:

[0023] The objective function and constraint conditions are:

[0024]

[0025] The objective function is expressed as minimizing the average dimming level of the luminaire and d is calculated using the following formula:

[0026]

[0027] where N is the total number of sensors, d i is the dimming level of the luminaire in the i-th area, is the average illuminance level set point, (E op , i) is the optimal illuminance value in the i-th area, is the dimming value at E m ; Equation (5) can be used as an analytical solution in the algorithm related to the dimming level of the LED lamp.

[0028] Furthermore, to minimize the average dimming level of the LED luminaire, several constraints need to be satisfied, which are divided into two categories of constraints: luminaire-based, i.e., the dimming ability of the luminaire, and illuminance-based, i.e., the average illuminance level E, the illuminance uniformity U0, and the illuminance value limit in the i-th control zone;

[0029] The dimming capacity limit of the luminaire is the minimum value D to which the luminaire can be dimmed min and the maximum value D max , and the typical value of the luminaire dimming capacity limit is in the range of 0 - 1, i.e., full dimming. The constraint is:

[0030] D min ≤d i ≤D max (6)

[0031] The average illuminance E is one of the indicators to measure human visual comfort, and it is recommended to maintain the average illuminance level E m at 500 lux. The constraint is:

[0032]

[0033] In addition to E, the illuminance uniformity U o can also be used as a measure of visual comfort. The minimum illuminance uniformity U o,min is 0.6. The constraint is:

[0034] U o ≥U o,min (8)

[0035] The limit of the illuminance E in the i-th control zone i is at least E i,min and at most E i,max . Depending on the different regions, their values are different. The constraint is:

[0036] E i,min ≤E i ≤E i,max (9)

[0037] Furthermore, the differential evolution optimization algorithm is used to solve the optical sensor layout problem, which specifically includes:

[0038] 1) Population initialization: Select an individual vector with a number of NP as the initial population, and these NP individual vectors are values in a D-dimensional continuous real-valued space. The i-th individual vector or target vector in the G-th generation is described using the following symbols:

[0039] X i，G =(X 1i，G ,X 2i，G …,X Di，G (10)

[0040] where \(G = 0, 1, \ldots, G\) max , \(G\) represents the generation to which the population belongs, \(G = 0\) represents the initialization of the population vector, and \(G\) max is the maximum generation, \(i\) represents the \(i\)-th individual vector, \(i = 1, 2, \ldots, NP\), \(D\) represents the \(D\)-dimensional space, and the differential evolution algorithm is to solve the global optimal solution in this \(D\)-dimensional continuous real-valued parameter space;

[0041] To ensure that the range of the initialized target vector can cover the entire solution space, the initialized target vector is expressed as the following formula:

[0042] \(X\) j，i = \(X\) j.min + \(rand\) j，i (0, 1)×(\(X\) j，max - \(X\) j，min ) (11)

[0043] where \(rand(0, 1)\) is a random number randomly generated in the interval (0, 1), and \(X\) min = {\(X\) 1，min , \(X\) 2，min , \(\ldots\), \(X\) D，min} represents the lower boundary of the target vector in the \(D\)-dimensional continuous real-valued space, and \(X\) max similarly represents the upper boundary of the target vector in the \(D\)-dimensional continuous real-valued space, and it needs to satisfy formula (8);

[0044] 2) Mutation operation: After step 1) is completed, the DE algorithm generates a mutant vector by adopting a mutation strategy for the initialized target vector \(X\) i，G , and the mutant vector is \(V\) i，G = (\(V\) 1i，G , \(V\) 2i，G , \(\ldots\), \(V\) Di，G ):

[0045] \(V\) i，G = \(X\) r1，G + \(F\)(\(X\) r2，G - \(X\) r3，G ) (12)

[0046] where \(r\) is a positive integer randomly selected from [1, NP], and the value range of \(F\) is [0, 1]. When local search is desired to achieve fast convergence, \(F\) should take a relatively small value;

[0047] 3) Crossover operation: Cross the parameters included in the mutant vector \(V\) i，G and the target vector \(X\) i，G to generate a new trial vector \(U\) i，G :

[0048]

[0049] where cr ∈ [0, 1] is the crossover probability;

[0050] 4) Selection operation: The trial vector U generated after mutation and crossover operations i，G , calculates the fitness value, and compares the calculated value with the target vector X i，G The calculated value for fitness calculation is compared, and the individual with a better result between the two is selected as the next-generation individual vector. The next-generation individual vector obtained by the DE algorithm is expressed as:

[0051]

[0052] where f is the fitness function, which is also the objective function (3).

[0053] The advantages and beneficial effects of the present invention are as follows:

[0054] The present invention proposes a method for optimizing the layout of photovoltaic sensors based on the differential evolution algorithm to minimize the initial cost and energy cost, and establishes the objective function and its constraints: the average dimming level, illuminance level, and uniformity of LED lamps. The differential evolution optimization algorithm (DE) is used to solve this problem. Finally, according to the results of the differential evolution algorithm, the optimal positions of the sensors are determined, thereby minimizing the lighting energy consumption to the greatest extent, maintaining the visual comfort of the building, and reducing the initial cost of the sensors at the same time.

[0055] A new method for the layout of the number and position of optical sensors is developed using the differential evolution algorithm to minimize costs (i.e., sensors and electrical energy), improve visual comfort, and controller performance (i.e., reduce complexity). This method has superior performance in terms of small computational effort and optimal solutions (i.e., the number and position of optical sensors). In addition, this method has strong practicability for the implementation of wired and wireless optical sensors in small and large buildings, new buildings, and renovation projects.

[0056] Equation (5) is newly defined in the present invention. The formula for calculating the dimming level value of the lamps in the i-th area has not been derived in previous studies because the value of d i is obtained from controller- and optimization-based solutions. For this reason, by considering maintaining the average illuminance level setpoint E m , the optimal illuminance value in the i-th area (E op ,i) and the dimming value d m at E Em to calculate d i .

[0057] To estimate the investment cost of the sensors and the capacity of the controller, it is crucial to determine the number of optical sensors to be installed in the room. In the general practice of lighting control strategies, the number of sensors should be less than or equal to the number of luminaires. In previous methods for optimizing the layout of sensors, control strategies were not incorporated. In terms of control strategies, pulse-width modulation (PWM) is usually used to control the dimming level. Under PWM control, it can be assumed that the dimming level of LED luminaires is linearly related to their output power. Reducing the dimming level of the luminaires to the minimum means reducing the electrical energy to the minimum.

[0058] The innovation lies in considering both the sensor location and the number of sensors simultaneously, taking the location and the number as the objectives of the optimization problem. By solving this optimization problem, the number of sensors is minimized while meeting the visual comfort requirements. Additionally, the illuminance level and the illuminance uniformity are used as constraint conditions. The more uniform the light distribution, the better the illuminance and the more comfortable the visual perception. The closer the illuminance uniformity is to 1, the better; conversely, the smaller it is, the more visual fatigue it causes. While achieving the economy of the optimization result, the light intensity can meet the requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 is a flowchart of the method for optimizing the layout of photovoltaic sensors based on the differential evolution algorithm provided by the present invention for the preferred embodiment. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0060] Next, the technical solutions in the embodiments of the present invention will be clearly and detailedly described in conjunction with the drawings in the embodiments of the present invention. The described embodiments are only a part of the embodiments of the present invention.

[0061] The technical solution of the present invention to solve the above technical problems is:

[0062] The overall framework flowchart of the method of the present invention is as Figure 1 shown. First, it is necessary to determine two illuminance-based metrics, which are the average illuminance level and the illuminance uniformity (U o ). To calculate several parameters need to be considered, including the reflection coefficients (i.e., the ceiling, walls, and floor), the parameters related to the luminaires, and the room area. To determine the in the measured area, it can be expressed by the following formula:

[0063]

[0064] where Nl is the number of luminaires, n l is the number of lamps in each luminaire, Φ nFor the output of the lamp, UF is the utilization factor, MF is the maintenance factor, BF is the ballast factor, and A is the surface area of the room. The common value of UF is 0.9. BF is always considered to be 1. In the design of lighting systems, the most commonly used value of MF is 0.8. To determine the value of UF, the room index RI needs to be calculated first. The formula for RI is expressed as follows:

[0065]

[0066] where L and W are the length and width of the room respectively, and H is the vertical distance from the luminaire to the working plane. After obtaining the RI value, the UF value can be determined by referring to the luminaire data sheet provided by the luminaire manufacturer. The illuminance uniformity (U o ) is one of the qualitative indicators related to the visual comfort of people in the room and can be calculated using the following formula:

[0067]

[0068] where E min is the minimum illuminance value of the room.

[0069] The optimal arrangement method of light sensors is crucial for determining the optimal number and location of light sensors to achieve energy conservation and satisfactory illuminance-based visual comfort. To implement the sensor arrangement method based on energy efficiency, the artificial lighting energy consumption is used as the objective function, and the illuminance indicators (i.e., illuminance level and uniformity) are used as the constraint conditions for optimization design.

[0070] To estimate the investment cost of the sensors and the capacity of the controller, it is crucial to determine the number of light sensors to be installed in the room. In the general practice of lighting control strategies, the number of sensors should be less than or equal to the number of luminaires. The main objective of the present invention is to find the optimal number and location of light sensors that meet energy conservation and visual comfort. To maximize the energy of the lighting system and the performance of the control system, LED luminaires are used. In terms of the control strategy, pulse width modulation (PWM) is used to control the dimming level. Under PWM control, it can be assumed that the dimming level of the LED luminaire is linearly related to its output power. Reducing the dimming degree of the luminaire to the minimum means reducing the electrical energy to the minimum. To find the optimal arrangement of light sensors corresponding to the minimum energy consumption, a problem formulation based on minimizing the dimming level of LED luminaires is established. The objective function can be expressed as minimizing the average dimming level of the illuminator as follows:

[0071]

[0072] Calculated using the following formula

[0073]

[0074] where N is the total number of sensors, d i is the dimming level of the i-th area luminaire, is the set point for maintaining the average illuminance level, (E op , i) is the optimal illuminance value for the i-th area, is the dimming value at E m . The value of (E op , i) is obtained from an optimization-based method (i.e., DE). This equation can be used as an analytical solution in an algorithm related to the dimming level of LED luminaires.

[0075] As mentioned above, the output power of LED luminaires is linearly related to their dimming degree. Therefore, the energy consumption of LED luminaires in a building can be expressed by the power demand (P D ), in kW. It can be calculated by considering the total power of the LED luminaires in the i-th area (P i ) and the calculated d i :

[0076]

[0077] where I is the total number of areas. To minimize the average dimming level of LED luminaires, several constraints need to be satisfied, which are divided into two categories of constraints: luminaire-based, i.e., the dimming ability of the luminaire, and illuminance-based, i.e., the average illuminance level the illuminance uniformity (U0) and illuminance value limit of the i-th control area. The dimming capacity limit of the luminaire is the minimum value (D min ) and the maximum value (D max ) to which the luminaire can be dimmed. Typical values of the luminaire dimming capacity limit are in the range of 0 - 1 (fully dimmed). The constraint is:

[0078] D min ≤d i ≤D max (6)

[0079] Average illuminance is one of the indicators to measure human visual comfort. It is recommended to maintain the average illuminance level (E m ) at 500 lux. The constraint is:

[0080]

[0081] In addition to the illuminance uniformity (U o ) can also be used as a measure of visual comfort. The minimum illuminance uniformity (U o,min ) is 0.6. The constraint is:

[0082] U o ≥Uo,min (8)

[0083] The illuminance E of the i-th control area i has a minimum limit of E i,min and a maximum of E i,max . Based on the different regions, their values are different. The constraint condition is:

[0084] E i,min ≤E i ≤E i,max (9)

[0085] In the case of considering the minimization of the average dimming level of LED lamps, the differential evolution algorithm for the optimal illuminance value of the i-th area is as follows:

[0086] 1) Population initialization: Select an individual vector with the number of NP as the initial population, and these NP individual vectors are values in the D-dimensional continuous real-value space. The i-th individual vector or target vector in the G-th generation is described by the following symbols:

[0087] X i，G =(X 1i，G , X 2i，G …, X Di，G ) (10)

[0088] In the formula, G = 0, 1…, G max , G represents the generation to which the population belongs, G = 0 represents the initialization of the population vector, G max is the maximum generation, i represents the i-th individual vector, i = 1, 2, …, NP, D represents the D-dimensional space, and the differential evolution algorithm is to solve the global optimal solution in this D-dimensional continuous real-value parameter space;

[0089] To ensure that the range of the initialized target vector can cover the entire solution space, the initialized target vector is expressed as the following formula:

[0090] X j，i =X j.min +rand j，i (0, 1)×(X j，max -X j，min ) (11)

[0091] In the formula, rand(0, 1) is a random number randomly generated in the interval (0, 1), X min ={X 1，min , X 2，min , …, X D，min} represents the lower boundary of the target vector in the D-dimensional continuous real-value space, and X max similarly represents the upper boundary of the target vector in the D-dimensional continuous real-value space, and it needs to satisfy formula (8);

[0092] 2) Mutation operation: After step 1) is completed, the DE algorithm generates a mutant vector by adopting a mutation strategy. The mutant vector is V i，G =(V i，G , V 1i，G …, V 2i，G …, V Di，G ):

[0093] V i，G =X r1，G +F(X r2，G -X r3，G ) (12)

[0094] Where r is a positive integer randomly selected from [1, NP], and the value range of F is [0, 1]. When local search is desired to achieve fast convergence, F should take a relatively small value;

[0095] 3) Crossover operation: Cross the parameters included in the mutant vector V i，G and the target vector X i，G to generate a new trial vector U i，G :

[0096]

[0097] Where cr ∈ [0, 1] is the crossover probability;

[0098] 4) Selection operation: Calculate the fitness value of the trial vector U i，G generated after mutation and crossover operations, and compare the calculated value with the fitness value calculated for the target vector X i，G . Select the individual with a better result between the two as the next-generation individual vector. The next-generation individual vector obtained by the DE algorithm is expressed as:

[0099]

[0100] where f is the fitness function, that is, the objective function (3).

[0101] The present invention utilizes a differential evolution optimization algorithm to optimize the lighting system control strategy by minimizing the dimming level of the lamps, thereby minimizing lighting energy consumption, maintaining the visual comfort of the building, and reducing the initial cost of the sensors at the same time.

[0102] It should also be noted that the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, commodity or device comprising a series of elements not only includes those elements but also includes other elements not expressly listed, or further includes elements inherent to such process, method, commodity or device. Without further limitation, an element defined by the statement "comprising one..." does not exclude the presence of additional identical elements in the process, method, commodity or device comprising said element.

[0103] The above embodiments should be understood as being only for illustrative purposes of the present invention and not for limiting the scope of protection of the present invention. After reading the content described in the present invention, those skilled in the art can make various changes or modifications to the present invention, and these equivalent changes and modifications also fall within the scope defined by the claims of the present invention.

Claims

1. A method for optimizing the layout of photovoltaic sensors based on the differential evolution algorithm, characterized in that, It includes the following steps: Determine the model of the average illuminance level and illuminance uniformity of the lighting lamp; take the energy consumption of lighting as the objective function and the illuminance level and uniformity as the constraint conditions for optimization design; use the differential evolution optimization algorithm to solve the photovoltaic sensor layout problem. The improvement of applying the differential evolution optimization algorithm to solve the photovoltaic sensor layout problem lies in being able to optimize the nonlinear function problem based on the photovoltaic sensor layout. Determine the optimal number of sensors based on the average dimming level and the area of the optimal illuminance value obtained by solving based on the differential evolution algorithm; The average illuminance level of the lighting lamp and the illuminance uniformity U o The model is as follows: where Nl is the number of luminaires, and n l is the number of lamps in each luminaire, Φ n is the output of the lamp, UF is the utilization factor, MF is the maintenance factor, BF is the ballast factor, and A is the surface area of the room; where E min is the minimum illuminance value of the room; Take the energy consumption of lighting as the objective function and the illuminance level and uniformity as the constraint conditions for optimization design, specifically including: The objective function and constraint conditions are: The objective function is expressed as minimizing the average dimming level of the lighting lamps and is calculated using the following formula where N is the total number of sensors, d i is the dimming level of the i-th area luminaire, is the set point for maintaining the average illuminance level, E op is the optimal illuminance value, is the dimming value at the illuminance E m At this point, Equation (5) can be used as an analytical solution for the dimming level of LED luminaires; To minimize the average dimming level of LED luminaires, several constraints need to be satisfied, which are divided into two categories of constraints: luminaire-based, i.e., the dimming ability of the luminaire, and illuminance-based, i.e., the average illuminance level The illuminance uniformity U0 and illuminance value limit of the i-th control area; The dimming capacity limit of the luminaire is the minimum value D to which the luminaire can be dimmed min and the maximum value D max , the typical value of the dimming capacity limit of the luminaire is in the range of 0 - 1, i.e., full dimming, and the constraint is: D min ≤ d i ≤ D max (6) Average illuminance is one of the indicators to measure the visual comfort of the human body, and it is recommended to maintain the average illuminance level at 500 lux, and the constraint conditions are: In addition to Illuminance uniformity U o can also be used as a measure of visual comfort. The minimum illuminance uniformity U o,min is 0.6, and the constraint condition is: U o ≥U o,min (8) The illuminance E of the i-th control area i has a minimum limit of E i,min and a maximum of E i,max and their values are different based on different regions, so the constraint condition is: E i,min ≤E i ≤E i,max (9); The use of the differential evolution optimization algorithm to solve the photovoltaic sensor layout problem specifically includes: 1) Population initialization: Select an individual vector with the number of NP as the initial population, and these NP individual vectors are values in the D-dimensional continuous real-value space. Use the following symbols to describe the i-th individual vector or target vector in the G-th generation: X i,G = (X 1i,G , X 2i,G …, X Di,G ) (10) where \(G = 0, 1, \ldots, G\) max , \(G\) represents the generation to which the population belongs, \(G = 0\) represents the initialization of the population vector, \(G\) max is the maximum generation, \(i\) represents the \(i\)-th individual vector, \(i = 1, 2, \ldots, NP\), \(D\) represents the \(D\)-dimensional space, and the differential evolution algorithm is to solve the global optimal solution in this \(D\)-dimensional continuous real-valued parameter space; To ensure that the range of the initialized target vector can cover the entire solution space, the initialized target vector is expressed as the following formula: X j,i = X j,min + rand j,i (0, 1) × (X j,max - X j,min ) (11) where rand(0, 1) is a random number randomly generated in the interval (0, 1), and X min = {X 1，min , X 2，min , …, X D，min} represents the lower boundary of the target vector in the D-dimensional continuous real-valued space, and X max Similarly, it is the upper boundary of the target vector in the D-dimensional continuous real-valued space, which needs to satisfy Equation (8); 2) Mutation operation: After step 1) is completed, the DE algorithm generates a mutant vector by adopting a mutation strategy. The mutant vector is V i，G =(V i，G , V i，G …, V 2i，G …, V Di，G ): V i，G = X r1，G + F(X r2，G - X r3，G )(12) In the formula, r is a positive integer randomly selected from [1, NP], and the value range of F is [0, 1]. When local search is desired to achieve fast convergence, F takes a smaller value; 3) Crossover operation: Cross the parameters included in the mutation vector V i，G and the target vector X i，G to generate a new trial vector U i，G : In the formula, cr ∈ [0, 1] is the crossover probability; 4) Selection operation: The trial vector U generated after mutation and crossover operations i，G , calculates the fitness value, and compares the calculated value with the target vector X i，G compares the fitness calculation values, and selects the individual with the better result between the two as the next-generation individual vector. The next-generation individual vector obtained by the DE algorithm is expressed as:

2. The method for optimizing the layout of a photovoltaic sensor based on a differential evolution algorithm according to claim 1, wherein To determine the value of UF, it is necessary to first calculate the room index RI. The formula for RI is expressed as follows: Where L and W are the length and width of the room respectively, and H is the vertical distance from the lamp to the working plane; after obtaining the RI value, the UF value can be determined by referring to the lamp data sheet provided by the lamp manufacturer.

3. A method for optimizing the layout of photovoltaic sensors based on the differential evolution algorithm according to claim 1, characterized in that, When the lamp uses an LED lamp, in terms of the control strategy, pulse width modulation (PWM) is used to control the dimming level. Under PWM control, it is assumed that the dimming level of the LED lamp is linearly related to its output power; reducing the dimming degree of the lamp to the lowest means reducing the electrical energy to the lowest; to find the optimal layout of the photovoltaic sensor corresponding to the minimum energy consumption, a problem formula based on minimizing the dimming level of the LED lamp is established.

Images