Multi-objective optimization method for photovoltaic sound barriers considering performance degradation over the entire life cycle
By constructing a full life cycle degradation prediction model and optimizing the design of photovoltaic sound barriers using the improved IA-NSGA-Ⅱ algorithm, the performance degradation problem of photovoltaic modules and sound-absorbing materials was solved, realizing the synergistic optimization of power generation and noise reduction of photovoltaic sound barriers and improving the overall efficiency throughout the entire life cycle.
Patent Information
- Application Number
- CN202611174468.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-04
- Publication Date
- 2026-11-17
AI Technical Summary
Existing photovoltaic sound barrier designs do not fully consider the attenuation of photovoltaic module output power over time and the performance degradation of sound-absorbing materials, resulting in a significant deviation between the overall efficiency throughout the entire life cycle and the initial design expectations. Furthermore, the coupling relationship between power generation and noise reduction systems has not been effectively addressed, making it difficult to achieve optimal synergy among multiple benefits.
A multi-objective optimization model is constructed, considering the life-cycle degradation of photovoltaic module output power and sound absorption coefficient of sound-absorbing material. The improved IA-NSGA-Ⅱ algorithm is used to solve the model. Combined with a coupling penalty mechanism and an adaptive collaborative convergence strategy, the design parameters of the photovoltaic sound barrier are optimized to achieve the optimal synergy between power generation, noise reduction and long-term economic efficiency.
It enables quantifiable prediction and synergistic optimization of the power generation efficiency and noise reduction performance of photovoltaic sound barriers throughout their entire life cycle, avoids design deviations caused by performance degradation, improves long-term comprehensive efficiency, and resolves the coupling conflict between power generation and noise reduction.
Smart Images

Figure CN123116780A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of photovoltaic sound barrier optimization design technology, specifically a multi-objective optimization method for photovoltaic sound barriers that considers the performance degradation throughout the entire life cycle. Background Technology
[0002] Photovoltaic sound barriers can simultaneously achieve clean power generation and noise suppression, representing an important technological direction for the decarbonization of transportation infrastructure and the improvement of the noise environment. Existing research on photovoltaic sound barriers primarily focuses on power generation efficiency and noise reduction performance under initial operating conditions, failing to fully consider the exponential decay of photovoltaic module output power over time and the performance degradation of sound-absorbing materials due to dust blockage and moisture erosion. This results in a significant deviation between the overall performance of the design scheme throughout its entire lifecycle and the initial design expectations.
[0003] Meanwhile, there is a significant strong coupling between the power generation system and the noise reduction system of photovoltaic sound barriers. Since sound barriers typically use porous sound-absorbing materials such as glass wool, the installation of photovoltaic modules will block the perforations on the surface of the sound-absorbing layer, reducing the sound wave incident channels and leading to a decrease in sound absorption efficiency. Photovoltaic modules themselves do not have sound absorption properties. If they are used directly as the sound barrier screen, the smooth glass surface will cause strong noise reflection, severely weakening the noise reduction effect. On the other hand, if the perforations of the sound-absorbing layer are too densely distributed when sound-absorbing materials are laid on photovoltaic modules, it will affect the installation area of the photovoltaic modules and reduce the photovoltaic power generation.
[0004] Existing optimization methods often treat power generation and noise reduction as independent objectives, failing to effectively address the coupling conflict between the two and making it difficult to achieve optimal synergy among all parties. Therefore, there is a need in this field for a multi-objective decision-making model that considers the entire life cycle of photovoltaic sound barriers to optimize power generation and noise reduction in order to achieve the optimal solution in terms of economy, environment, and noise reduction. Summary of the Invention
[0005] To address or partially address the problems existing in related technologies, this invention provides a multi-objective optimization method for photovoltaic sound barriers that considers performance degradation throughout the entire life cycle.
[0006] To achieve the above objectives, the technical solution of the present invention is as follows: A multi-objective optimization method for photovoltaic sound barriers considering performance degradation throughout their entire life cycle is provided, characterized by comprising the following steps: S1. Collect basic parameters, including photovoltaic module parameters, sound absorption and insulation structure parameters, support structure parameters, and environmental parameters of the installation site; S2. Based on the basic parameters obtained in S1, the design parameters to be optimized are combined into a decision variable vector, and a multi-objective optimization model including photovoltaic power generation efficiency, noise reduction performance and life cycle cost is constructed, and the constraints of the multi-objective optimization model are set. S3. Introduce the time dimension and construct a full life cycle attenuation prediction model that includes photovoltaic module output power attenuation, sound absorption material sound absorption coefficient attenuation and coupling attenuation factor attenuation. Embed the full life cycle attenuation prediction model into the multi-objective optimization model constructed in S2 to obtain a multi-objective optimization model with embedded performance attenuation characteristics. S4. The improved IA-NSGA-Ⅱ algorithm, which includes a penalty mechanism and an adaptive collaborative convergence strategy, is used to solve the multi-objective optimization model of the S3 embedded full life cycle decay prediction model, and the Pareto optimal solution set that takes into account the synergistic optimization of power generation, noise reduction and long-term economic efficiency is obtained. S5. Select the optimal design scheme from the Pareto optimal solution set output by S4, and carry out the structural design, component and construction installation of the photovoltaic sound barrier according to the design scheme; S6. After the photovoltaic sound barrier is put into actual operation, collect on-site operation data regularly, compare it with the output of the full life cycle attenuation prediction model, correct the model parameters online according to the deviation, and feed the corrected parameters back to the attenuation prediction model for updating subsequent predictions.
[0007] Furthermore, the basic parameters in S1 specifically include: Photovoltaic module parameters include photovoltaic module conversion efficiency, photovoltaic module weight per unit area, photovoltaic module tilt angle, photovoltaic module installation area, photovoltaic module cleaning cycle, and tilt angle adjustment frequency; The structural parameters for sound absorption and insulation include the thickness of the sound-absorbing layer, the thickness of the sound-insulating layer, the perforation rate of the sound-absorbing material, the sound absorption coefficient of the sound-absorbing material, the density of the sound-insulating material, and the weight per unit volume of the sound-absorbing material. Among them, the sound-absorbing layer is a porous sound-absorbing material layer (such as glass wool or rock wool) set on the side facing the noise source, which is used to absorb sound wave energy; the sound-insulating layer is a dense barrier layer (such as metal plate or composite board) set on the side away from the noise source, which is used to block sound wave transmission.
[0008] Support structure parameters include the maximum load-bearing capacity of the support structure; Environmental parameters for the installation site include annual average solar irradiance, air characteristic impedance, and the speed of sound in the air.
[0009] Furthermore, the specific process of S2 is as follows: S2.1 Determine the decision variable vector X: The design parameters to be optimized are combined into an 8-dimensional decision variable vector. X , represented as: In the formula, α Photovoltaic module tilt angle; S This refers to the actual area where the photovoltaic modules are installed. η p For component conversion efficiency; d1 represents the thickness of the sound-absorbing layer; d 2 represents the thickness of the sound insulation layer; The perforation rate of the sound-absorbing material; T For the cleaning cycle of photovoltaic modules; f Adjust the frequency to adjust the tilt angle.
[0010] S2.2 Constructing a multi-objective optimization function Three interrelated and mutually restrictive optimization objectives are set to maximize power generation efficiency, maximize noise reduction performance, and minimize total life cycle cost, respectively. These three objectives are strongly coupled.
[0011] S2.2.1 Calculate photovoltaic power generation efficiency Based on the annual power generation of photovoltaic systems E year To quantify the indicator, the calculation comprehensively considers the annual average radiation. H avg The formulas for system efficiency, temperature degradation, and power index degradation of photovoltaic modules over their entire lifespan are as follows: In the formula, γ is the overall efficiency coefficient of the photovoltaic system; δ t This is the temperature decay coefficient; Y For cumulative years; The annual degradation rate of photovoltaic modules; This indicates the amount of electricity generated in the first year; The equivalent number of years of decay is equivalent to converting the annual decay into several full-power years; the whole formula is equivalent to the power generation in the first year multiplied by the equivalent number of full-power years = the total power generation over the entire life cycle.
[0012] S2.2.2 Calculation of noise reduction performance Noise reduction performance is measured by the average annual noise reduction amount. N avg As a quantitative indicator, the calculation formula is: in, N abs This refers to the noise reduction amount of the sound-absorbing layer; N ins This represents the noise reduction amount of the sound insulation layer; Δ N Noise reduction loss due to shading of photovoltaic modules; This is the attenuation term of the sound-absorbing layer performance index. These are the preset equivalent acoustic parameters or static correction coefficients related to the geometric configuration of the sound-absorbing structure; For the occlusion loss growth term; γ d The annual growth rate of the loss; t For time (years); The calculation formulas for each component in the noise reduction performance formula are as follows: In the formula, α abs The sound absorption coefficient of the sound-absorbing material; ρ Density of sound insulation material; c The speed of sound in air; Z 0 represents the characteristic impedance of air. k This is the area influence coefficient; S total This represents the total usable surface area of the sound barrier.
[0013] S2.2.3 Calculate the total life cycle cost Total life cycle cost C LCC The calculation formula is: in, C init For initial investment cost, C op For operation and maintenance costs, C en Energy consumption cost.
[0014] S2.3 Setting Constraints S2.3.1, The boundary constraints of the decision variables are: S2.3.2, Performance constraints are: S2.3.3, Structural safety constraints are: In the formula, ρ p This refers to the weight per unit area of the photovoltaic module. ρ abs The weight per unit volume of the sound-absorbing material; W max To support the maximum load-bearing capacity of the structure.
[0015] S2.3.4 Coupling and Coordination Constraints To quantify and suppress the power generation-noise reduction coupling conflict, a coupling conflict coefficient is introduced. The calculation formula is: In the formula, P ( α () represents the actual power generation. N ( α The tilt angle of the photovoltaic module is... α The actual noise reduction amount at that time; P 0 represents the ideal power generation capacity of a photovoltaic module; N 0 represents the ideal noise reduction level for the noise reduction system. P 0 represents the power generation of a photovoltaic module under ideal boundary conditions. It usually refers to the theoretical maximum output power when the module is at its optimal tilt angle and there is no shading, no attenuation, and no dust cover under standard test conditions. P The 0 value is generally calculated by combining the nominal power of the component with the solar irradiance correction factor under the local optimal tilt angle, or obtained through short-cycle peak testing in a controlled laboratory environment, and serves as a normalized benchmark for measuring actual operating power. N 0 represents the theoretical maximum noise reduction of the noise reduction system under ideal conditions where there is no photovoltaic module obstruction, no material aging, and no dust blockage, and the sound-absorbing and sound-insulating structure is fully covered on the surface of the sound barrier according to the designed thickness and perforation rate. N The 0 value is usually calculated based on physical properties such as sound absorption coefficient, density of sound insulation material, and air characteristic impedance, using a noise reduction performance formula under conditions of no obstruction loss.
[0016] Coupling conflict coefficient The constraints that need to be satisfied are: .
[0017] S2.4 Constructing a multi-objective optimization model The strongly coupled multi-objective optimization model of photovoltaic sound barrier power generation and noise reduction is as follows: Furthermore, the full lifecycle degradation prediction model constructed by S3 is as follows: The output power of photovoltaic modules decreases exponentially over time, as shown in the formula: in, P ( t ) No. t Actual annual output power; parameters β p The optimal annual return is 0.5%, which is a typical value for monocrystalline PERC modules.
[0018] The attenuation of the sound absorption coefficient of sound-absorbing materials is affected by dust blockage and moisture erosion, as shown in the formula: In the formula, For the first t The actual sound absorption coefficient for the year; The initial sound absorption coefficient; β s The annual attenuation rate of the sound absorption coefficient of the sound-absorbing material. β s Preferred rate: 0.3% / year; α res This represents the residual sound absorption coefficient.
[0019] The two attenuation methods are coupled through the attenuation factor. ξ ( t Related: In the formula, The coupling attenuation coefficient is initially set to 0.2, reflecting the feedback effect that leads to a decrease in clean frequency due to a decline in power generation efficiency, which in turn accelerates the degradation of noise reduction performance.
[0020] The lifecycle degradation prediction model is embedded into the multi-objective optimization model constructed by S2, that is, the instantaneous performance target originally based on a single year or static operating condition is reconstructed into a time-varying cumulative target across the entire lifecycle. , These objectives are respectively substituted into the power generation efficiency and noise reduction performance objectives of the S2 multi-objective optimization model. Simultaneously, the life-cycle cost objective, based on the original initial investment, operation and maintenance, and energy consumption costs, further internalizes the additional cleaning maintenance expenses and power generation revenue loss caused by performance degradation. Based on this, by coupling the degradation factor, the independent degradation processes of photovoltaics and noise reduction are linked into a mutually feeding degradation process. This causes the decline in power generation efficiency to accelerate the noise reduction performance degradation of the sound-absorbing and insulating structure inversely through the mediating variable of reducing the cleaning frequency. Thus, the three objectives of power generation, noise reduction, and cost are transformed from a static, decoupled state into a dynamic, strongly coupled, and time-varying synergistic optimization system.
[0021] Furthermore, the process of solving the multi-objective optimization model of the S3 embedded life-cycle decay prediction model using the improved IA-NSGA-Ⅱ algorithm in S4 is as follows: S4.1 Real Number Encoding: Each individual is encoded as an 8-dimensional real number vector, which is linked to the decision variable vector. X =[ α , S , η p , d 1, d 2, , T , f ] T One-to-one correspondence, the value range of each dimension is consistent with the boundary constraints of the decision variables; S4.2 Population Initialization: Several individuals are randomly generated within the feature space, each representing a photovoltaic sound barrier design scheme. Specifically, during initialization, the algorithm, within a multidimensional feature space composed of eight decision variables, assigns values to each individual in the population dimension-by-dimensionally using a random number generator, based on the preset upper and lower bound constraints of each dimension. This uniformly or randomly distributes several initial design schemes within the solution space, forming the initial population. The boundary constraints of the decision variables are shown in S2.3.1.
[0022] S4.3 Introduce a coupling penalty mechanism: Set the coupling penalty factor ω to: In the formula, k p This is the penalty coefficient; λ This represents the coupling conflict coefficient.
[0023] The coupling penalty factor is incorporated into the following three objective functions for modification, resulting in the following fitness function: (1) Power generation efficiency fitness function for: (2) Noise reduction performance fitness function for: (3) Cost fitness function (Transforming the cost minimization problem into a cost maximization problem): In the formula, F 1. F 2. F 3 represents the corrected fitness function value; C LCC,max This represents the maximum value of the total lifecycle cost.
[0024] S4.4 Design an adaptive collaborative convergence strategy To improve the convergence speed and optimization accuracy of the algorithm, an adaptive collaborative convergence strategy is designed to dynamically adjust the crossover probability. With the probability of mutation This allows the algorithm to quickly search for the global optimum in the early stages of iteration and to converge accurately in the later stages of iteration, avoiding getting trapped in local optima.
[0025] (1) Adaptive crossover probability The dynamic adjustment formula is as follows: In the formula, p c For adaptive crossover probability;p c0 The initial crossover probability; k c This is the crossover probability adjustment coefficient; n This represents the current iteration step number; G max This represents the maximum number of iterations.
[0026] (2) Adaptive mutation probability The dynamic adjustment formula is as follows: The symbols in the formula are defined as follows: For adaptive mutation probability; p m0 The initial mutation probability; k m This is the adjustment coefficient for the probability of variation.
[0027] S4.5 Population Evolution: Based on the corrected fitness function values, perform non-dominated sorting and crowding calculations on the individuals in the population, and sequentially execute binary tournament selection to choose parent individuals; when the generated random number for each pair of parents is less than or equal to the adaptive crossover probability... Then perform simulated binary crossover; otherwise, directly copy. When the generated random number for each dimension of each offspring is less than or equal to the adaptive mutation probability... Then, a polynomial mutation is performed to generate a new generation of population.
[0028] S4.6 Repeat the iteration until the maximum number of iterations G is reached. max Output the Pareto optimal solution set.
[0029] Furthermore, the online correction model parameters in S6 include the photovoltaic module degradation rate. β p Attenuation rate of sound-absorbing materials β s and coupling attenuation coefficient The specific correction method is as follows: For photovoltaic module degradation rate β p Using periodically collected actual power generation data, based on the formula The least squares method or exponential regression is used for fitting, and the results are dynamically updated. β p The measured estimated value; Annual attenuation rate of sound absorption coefficient of sound-absorbing materials β s Based on periodic measured data of sound absorption coefficient or noise reduction behind the barrier, and using the formula... Perform exponential fitting, after subtracting the residual sound absorption coefficient α resRe-identification β s ; For coupling attenuation coefficient Based on coupling attenuation factor By using periodically collected synchronous observation data of power generation and noise reduction, As the dependent variable, Perform linear regression on the independent variable and re-estimate the regression coefficients. value.
[0030] This invention provides a multi-objective optimization method for photovoltaic sound barriers that considers performance degradation over the entire life cycle, and has the following beneficial effects: (1) This invention solves the technical problem in the prior art that only focuses on the initial operating condition design and does not fully consider the long-term performance degradation, resulting in a serious deviation between the overall performance and the initial design expectation. This is achieved by constructing a full life cycle degradation prediction model for the output power of embedded photovoltaic modules, the sound absorption coefficient of sound-absorbing materials, and the coupling degradation factor. As a result, the power generation efficiency and noise reduction performance degradation law of the design scheme can be quantified and predicted throughout the entire life cycle, ensuring that the long-term overall performance is highly consistent with the design expectation.
[0031] (2) This invention quantifies the degree of coupling conflict between power generation and noise reduction by introducing a coupling conflict coefficient, and incorporates this coefficient into the coupling penalty mechanism of the constraint conditions and the IA-NSGA-II algorithm. This solves the technical problem in the prior art that power generation and noise reduction are optimized as independent objectives and fail to effectively handle the strong coupling conflict between the two. It achieves the synergistic optimality of power generation efficiency and noise reduction performance under strong coupling constraints, and avoids one party excessively sacrificing the other.
[0032] (3) This invention designs an adaptive collaborative convergence strategy that decreases the crossover probability and increases the mutation probability as the iteration progresses. It also introduces a coupling penalty factor to penalize individuals that violate the coupling coordination constraints. This solves the technical problem of slow convergence speed and susceptibility to local optima caused by the fixed crossover and mutation probabilities in the conventional NSGA-II algorithm. As a result, it achieves significantly higher optimization accuracy and shorter convergence steps compared to the conventional NSGA-II algorithm. Attached Figure Description
[0033] Figure 1 This is a flowchart of the steps of the present invention.
[0034] Figure 2 This is a flowchart of the IA-NSGA-Ⅱ algorithm of the present invention.
[0035] Figure 3 This is a graph comparing the convergence speed and overall optimization performance of each algorithm in Embodiment 1 of the present invention.
[0036] Figure 4 This is a diagram showing the change in dynamic optimization accuracy during the iteration process of each algorithm in Embodiment 1 of the present invention.
[0037] Figure 5 This is a comparison chart of the monthly dynamic changes in performance of various schemes under different operating conditions in Embodiment 1 of the present invention. Detailed Implementation
[0038] The present invention will be further described in detail below with reference to specific embodiments and accompanying drawings, but the scope of protection of the present invention is not limited to the content described.
[0039] This invention provides a multi-objective optimization method for photovoltaic sound barriers that considers performance degradation over the entire life cycle, such as... Figure 1 As shown, it includes the following steps: S1. Collect basic parameters The collected parameters include: photovoltaic module parameters, sound absorption and insulation structure parameters, supporting structure parameters, and environmental parameters of the installation site; specifically: Photovoltaic module parameters include photovoltaic module conversion efficiency, photovoltaic module weight per unit area, photovoltaic module tilt angle, photovoltaic module installation area, photovoltaic module cleaning cycle, and tilt angle adjustment frequency; The structural parameters for sound absorption and insulation include the thickness of the sound-absorbing layer, the thickness of the sound-insulating layer, the perforation rate of the sound-absorbing material, the sound absorption coefficient of the sound-absorbing material, the density of the sound-insulating material, and the weight per unit volume of the sound-absorbing material. Among them, the sound-absorbing layer is a porous sound-absorbing material layer (such as glass wool or rock wool) set on the side facing the noise source, which is used to absorb sound wave energy; the sound-insulating layer is a dense barrier layer (such as metal plate or composite board) set on the side away from the noise source, which is used to block sound wave transmission.
[0040] Support structure parameters include the maximum load-bearing capacity of the support structure; Environmental parameters for the installation site include annual average solar irradiance, air characteristic impedance, and the speed of sound in the air.
[0041] S2. Construct a multi-objective optimization model for strong coupling of photovoltaic sound barrier power generation and noise reduction. Based on the basic parameters obtained from S1, a mathematical multi-objective optimization model is established that can accurately describe the strong coupling and conflict relationship between power generation and noise reduction, providing an objective and constraint framework for subsequent solutions. The multi-objective optimization model includes decision variables, objective functions, and constraints.
[0042] S2.1 Determine the decision variable vector X: The design parameters to be optimized are combined into an 8-dimensional decision variable vector. X , represented as: In the formula, α Photovoltaic module tilt angle; SThis refers to the actual area where the photovoltaic modules are installed. η p For component conversion efficiency; d 1 represents the thickness of the sound-absorbing layer; d 2 represents the thickness of the sound insulation layer; The perforation rate of the sound-absorbing material; T For the cleaning cycle of photovoltaic modules; f Adjust the frequency to adjust the tilt angle.
[0043] S2.2 Constructing a multi-objective optimization function Three interrelated and mutually restrictive optimization objectives are set to maximize power generation efficiency, maximize noise reduction performance, and minimize total life cycle cost, respectively. These three objectives are strongly coupled.
[0044] S2.2.1 Calculate photovoltaic power generation efficiency Based on the annual power generation of photovoltaic systems E year To quantify the indicator, the calculation comprehensively considers the annual average radiation. H avg The formulas for system efficiency, temperature degradation, and power index degradation of photovoltaic modules over their entire lifespan are as follows: In the formula, γ is the overall efficiency coefficient of the photovoltaic system; δ t This is the temperature decay coefficient; Y For cumulative years; The annual degradation rate of photovoltaic modules; This indicates the amount of electricity generated in the first year; The equivalent number of years of decay is equivalent to converting the annual decay into several full-power years; the whole formula is equivalent to the power generation in the first year multiplied by the equivalent number of full-power years = the total power generation over the entire life cycle.
[0045] S2.2.2 Calculation of noise reduction performance Noise reduction performance is measured by the average annual noise reduction amount. N avg As a quantitative indicator, the calculation formula is: in, N abs This refers to the noise reduction amount of the sound-absorbing layer; N ins This represents the noise reduction amount of the sound insulation layer; Δ N Noise reduction loss due to shading of photovoltaic modules; This is the attenuation term of the sound-absorbing layer performance index. It is a preset equivalent acoustic parameter or static correction coefficient related to the geometric configuration of the sound-absorbing structure, used to calculate the noise reduction of the sound-absorbing layer in the initial design state, and is a static design parameter. For the occlusion loss growth term; γ d The annual growth rate of the loss; t For time (years).
[0046] The calculation formulas for each component in the noise reduction performance formula are as follows: In the formula, α abs The sound absorption coefficient of the sound-absorbing material; ρ Density of sound insulation material; c The speed of sound in air; Z 0 represents the characteristic impedance of air; k This is the area influence coefficient; S total This represents the total usable surface area of the sound barrier.
[0047] S2.2.3 Calculate the total life cycle cost Total life cycle cost C LCC The calculation formula is: in, C init For initial investment cost, C op For operation and maintenance costs, C en Energy consumption cost.
[0048] S2.3 Setting Constraints S2.3.1, The boundary constraints of the decision variables are: S2.3.2, Performance constraints are: S2.3.3, Structural safety constraints are: In the formula, ρ p This refers to the weight per unit area of the photovoltaic module. S total This represents the total usable surface area of the sound barrier. ρ absThe weight per unit volume of the sound-absorbing material; W max To support the maximum load-bearing capacity of the structure.
[0049] S2.3.4 Coupling and Coordination Constraints To quantify and suppress the power generation-noise reduction coupling conflict, a coupling conflict coefficient is introduced. The calculation formula is: In the formula, P ( α () represents the actual power generation. N ( α The tilt angle of the photovoltaic module is... α The actual noise reduction amount at that time; P 0 represents the ideal power generation capacity of a photovoltaic module; N 0 represents the ideal noise reduction level for the noise reduction system.
[0050] Coupling conflict coefficient The constraints that need to be satisfied are: .
[0051] S2.4 Constructing a multi-objective optimization model The strongly coupled multi-objective optimization model of photovoltaic sound barrier power generation and noise reduction is as follows: S3. Construction and Embedding of the Full Life Cycle Decay Prediction Model Based on the static multi-objective optimization model constructed in S2, to accurately describe the performance degradation of photovoltaic modules and sound-absorbing structures during long-term operation, a time dimension is introduced to construct a full life-cycle degradation prediction model. This model predicts the performance degradation of photovoltaic modules and sound-absorbing structures during long-term operation. The full life-cycle degradation prediction model includes the output power degradation of photovoltaic modules, the sound absorption coefficient degradation of sound-absorbing materials, and the coupling degradation factor, as detailed below: The output power of photovoltaic modules decreases exponentially over time, as shown in the formula: in, P ( t ) No. t Actual annual output power; parameters β p The optimal annual return is 0.5%, which is a typical value for monocrystalline PERC modules.
[0052] The attenuation of the sound absorption coefficient of sound-absorbing materials is affected by dust blockage and moisture erosion, as shown in the formula: In the formula, For the first tThe actual sound absorption coefficient for the year; The initial sound absorption coefficient; β s The annual attenuation rate of the sound absorption coefficient of the sound-absorbing material is used in the whole life cycle attenuation prediction model and is a dynamic degradation parameter. β s Preferred rate: 0.3% / year; α res This represents the residual sound absorption coefficient.
[0053] The two attenuation methods are coupled through the attenuation factor. ξ ( t Related: In the formula, This is the coupling attenuation coefficient, initially set to 0.2. Its function is to correlate the dynamic feedback effect between the attenuation of the power generation system and the attenuation of the noise reduction system. It belongs to the coupling parameter at the level of performance degradation throughout the entire life cycle and is used to construct the coupling attenuation factor. It is a dimensionless time-varying decay correlation parameter.
[0054] The full life cycle degradation prediction model is embedded into the multi-objective optimization model constructed by S2, and the original static objective is replaced with the cumulative power generation efficiency, cumulative noise reduction performance and full life cycle cost as optimization objectives.
[0055] The full lifecycle degradation prediction model is embedded into the multi-objective optimization model constructed by S2, that is, the instantaneous performance target originally based on a single year or static operating conditions is reconstructed into a time-varying cumulative target across the entire lifecycle. Specifically, the photovoltaic module output power exhibits an exponential decay over time ( Substitute into S2 E year The integral expression shows that the annual power generation decreases with increasing years of operation; the degradation law of the sound absorption coefficient of sound-absorbing materials is affected by dust blockage, moisture erosion, etc. Substitute into S2 N avg The integral expression makes the noise reduction contribution decrease with the number of years of operation. Meanwhile, the life-cycle cost target further internalizes the additional cleaning maintenance expenses and power generation revenue loss caused by performance degradation, on top of the original initial investment, operation and maintenance, and energy consumption costs. Based on this, by coupling the degradation factor ( By linking the independent degradation processes of the photovoltaic system and the noise reduction system into a mutually feeding degradation process, the decline in power generation efficiency is reversed by reducing the cleaning frequency as an intermediary variable, thus accelerating the degradation of the noise reduction performance of the sound-absorbing and insulating structure. This transforms the three objectives of power generation, noise reduction, and cost from a static and decoupled state into a dynamic, strongly coupled, and time-varying synergistic optimization system. This optimization algorithm no longer only pursues the peak power generation and noise reduction during the initial stage of system operation, but can predict the performance evolution trajectory throughout the entire life cycle, reserving degradation margins in advance in component layout, structural parameters, and operation and maintenance strategies.
[0056] S4. Design an improved IA-NSGA-II algorithm The multi-objective optimization model of the S3 embedded full life cycle decay prediction model is solved, and the flow field of the improved IA-NSGA-Ⅱ algorithm is as follows: Figure 2 As shown, the details are as follows: S4.1 Real Number Encoding: Each individual is encoded as an 8-dimensional real number vector, which is linked to the decision variable vector. X =[ α , S , η p , d 1, d 2, , T , f ] T One-to-one correspondence, the value range of each dimension is consistent with the boundary constraints of the decision variables; S4.2 Population Initialization: Several individuals are randomly generated within the feature space, each representing a photovoltaic sound barrier design scheme. Specifically, during initialization, the algorithm, within a multidimensional feature space composed of eight decision variables, assigns values to each individual in the population dimension-by-dimensionally using a random number generator, based on the preset upper and lower bound constraints of each dimension. This uniformly or randomly distributes several initial design schemes within the solution space, forming the initial population. The boundary constraints of the decision variables are shown in S2.3.1.
[0057] S4.3 Introduce a coupling penalty mechanism: Set the coupling penalty factor ω to: In the formula, k p This is the penalty coefficient; λ This represents the coupling conflict coefficient.
[0058] The coupling penalty factor is incorporated into the following three objective functions for modification, resulting in the following fitness function: (1) Power generation efficiency fitness function for: (2) Noise reduction performance fitness function for: (3) Cost fitness function (Transforming the cost minimization problem into a cost maximization problem): In the formula, F 1. F 2. F 3 represents the corrected fitness function value; C LCC,max This represents the maximum value of the total lifecycle cost.
[0059] S4.4 Design an adaptive collaborative convergence strategy To improve the convergence speed and optimization accuracy of the algorithm, an adaptive collaborative convergence strategy is designed to dynamically adjust the crossover probability. With the probability of mutation This allows the algorithm to quickly search for the global optimum in the early stages of iteration and to converge accurately in the later stages of iteration, avoiding getting trapped in local optima.
[0060] (1) Adaptive crossover probability The dynamic adjustment formula is as follows: In the formula, p c For adaptive crossover probability; p c0 The initial crossover probability; k c This is the crossover probability adjustment coefficient; n This represents the current iteration step number; G max This represents the maximum number of iterations.
[0061] (2) Adaptive mutation probability The dynamic adjustment formula is as follows: The symbols in the formula are defined as follows: For adaptive mutation probability; p m0 The initial mutation probability; k m This is the adjustment coefficient for the probability of variation.
[0062] S4.5 Population Evolution: Based on the corrected fitness function values, perform non-dominated sorting and crowding calculations on the individuals in the population, and sequentially execute binary tournament selection to choose parent individuals; when the generated random number for each pair of parents is less than or equal to the adaptive crossover probability... Then perform simulated binary crossover; otherwise, directly copy. When the generated random number for each dimension of each offspring is less than or equal to the adaptive mutation probability... Then, a polynomial mutation is performed to generate a new generation of population.
[0063] S4.6 Repeat the iteration until the maximum number of iterations G is reached. max Output the Pareto optimal solution set.
[0064] S5. Select the optimal solution and implement it. The optimal design scheme is selected from the Pareto optimal solution set output by S4 (either arbitrarily or based on engineering preferences). The selected optimal design scheme explicitly provides the decision variables. α , S , η p , d 1, d 2, , T , f The optimal combination of values is determined, and the structural design, component design, and construction and installation of the photovoltaic sound barrier are carried out according to this design scheme.
[0065] S6, Feedback Correction After the photovoltaic sound barrier is put into actual operation, a closed-loop optimization mechanism is established to regularly collect data on the actual power generation and noise reduction on site, compare the output of the S3 full life cycle attenuation prediction model, and use the measured data to optimize the key parameters in the model. β p , β s and k Online corrections are performed to make the prediction model more closely match actual working conditions. The corrected parameters are fed back into the full life cycle degradation prediction model in real time for subsequent optimization iterations, thus forming a complete closed loop of "design-implementation-monitoring-correction-re-optimization".
[0066] The specific online correction process is as follows: Relying on the physical test platform and its synchronously configured high-precision environmental monitoring system and performance monitoring system, this invention establishes a rolling data-driven parameter identification and model update mechanism based on 132,000 sets of valid data accumulated over 18 months of actual measurements. Regarding the photovoltaic module degradation rate... β p Using actual power generation data collected periodically (e.g., monthly), based on the formula The least squares method or exponential regression is used for fitting, and the results are dynamically updated. β p The measured estimated value; the annual attenuation rate of the sound absorption coefficient of the sound-absorbing material. β sBased on periodic measured data of sound absorption coefficient or noise reduction behind the barrier, and using the formula... Perform exponential fitting, after subtracting the residual sound absorption coefficient α res Re-identification β s For coupling attenuation coefficient Based on coupling attenuation factor By using periodically collected synchronous observation data of power generation and noise reduction, As the dependent variable, Perform linear regression on the independent variable and re-estimate the regression coefficients. value.
[0067] The above online correction process employs a rolling time window mechanism, re-executing parameter identification using newly added measured data at regular intervals, and updating the parameters accordingly. β p , β s and The full life cycle degradation prediction model is embedded in real time, and then the optimality of the current decision variables is re-evaluated through the IA-NSGA-II algorithm to achieve continuous alignment between the degradation prediction model and the actual working conditions.
[0068] To further analyze the coupling relationship between the photovoltaic module installation area and noise reduction performance, this invention introduces an installation area influence coefficient. The calculation formula is as follows: In the formula, k This is the area influence coefficient; N ( S The area covered by photovoltaic modules is The actual noise reduction amount at that time.
[0069] An impact coefficient η on the installation area is introduced to describe the coupling conflict between the photovoltaic module installation area and noise reduction performance. η represents the actual installation area of the photovoltaic module. S Total usable area of the sound barrier S total The ratio, through the area influence coefficient k Mapped to noise reduction N ( S (On) Laying area S Directly incorporated into the 8-dimensional decision variable vector X When the optimizer searches for the optimal solution, it performs each step... S All adjustments must involve the noise reduction loss represented by η. The coupling relationship defined by η is actually embedded in the noise reduction performance fitness function and coupling coordination constraints, so that the algorithm finds a Pareto optimal balance between increasing power generation by laying more photovoltaic panels and maintaining noise reduction by laying fewer photovoltaic panels.
[0070] k This is the area impact coefficient, which quantifies the area of photovoltaic module installation. S For actual noise reduction N ( S The direct impact of the sound barrier is a coupling parameter at the structural layout level. This coefficient is used to correct for noise reduction losses caused by the physical obstruction of the sound-absorbing layer surface by photovoltaic modules and the reduction of sound wave incident channels. It reflects the static geometric coupling relationship between space occupancy and acoustic performance, and its value is related to the total usable area of the sound barrier surface. S total The effective sound absorption area of the perforated panel is related to the component arrangement and is a preset value.
[0071] Based on the above technical solution, the present invention will be further illustrated by the following embodiments: Example 1 I. Experimental Platform and Testing Conditions In this embodiment, based on the method of the present invention, a physical test platform was built and a full life-cycle field test study was conducted over an 18-month period. Through multi-dimensional comparison, mechanism tracing, and quantitative analysis, all measured data were repeated three times. The test site was located on a core section of a highway with an average daily traffic flow of 12,000 vehicles. The noise source was mainly mid-to-high frequency traffic noise, with an average annual irradiance of 4.2 kWh / m². The test covered typical service conditions such as high temperature, heavy rain, and sandstorms, which could fully simulate the actual operating environment of the photovoltaic sound barrier. The natural environmental conditions changed continuously during the test, and extreme weather was not artificially arranged in segments. The test platform was 50m long and 4.5m high, using PERC bifacial photovoltaic modules and a composite sound absorption and insulation structure to ensure the engineering representativeness of the test carrier. A high-precision environmental monitoring system and an efficiency monitoring system were built simultaneously, accumulating 132,000 sets of valid data. This embodiment was implemented under the above conditions using the technical solution of the present invention.
[0072] II. Algorithm Performance Verification To verify the effectiveness of the algorithm of this invention, comparative experiments were conducted under the same conditions with four commonly used algorithms—NSGA-II, MOPSO, NSGA-III, and MOEA / D—as well as the innovative mechanism ablation algorithm. The innovative mechanism ablation algorithm is a variant of the invented IA-NSGA-II algorithm, which removes the coupling penalty mechanism and adaptive convergence strategy, and is used to separately verify the effectiveness of these two innovative mechanisms. The comparison results are as follows: Figure 3 and 4 As shown.
[0073] Figure 3This diagram illustrates a performance comparison of the method of this invention in Embodiment 1, along with four commonly used algorithms—NSGA-II, MOPSO, NSGA-III, and MOEA / D—and an innovative ablation mechanism algorithm. Figure 3 As can be seen, the improved IA-NSGA-II algorithm of this invention has the fastest convergence speed, reaching the convergence condition in 120 iterations, which is significantly better than other algorithms. Its comprehensive optimization performance curve is always at the highest level, with fluctuation range ≤0.45%, showing extremely strong stability. The conventional NSGA-II algorithm has the slowest convergence and the worst comprehensive optimization performance. The core reason is that this algorithm is prone to getting trapped in local optima, resulting in a significant decrease in optimization efficiency and accuracy. The MOPSO algorithm has slightly better global optimization ability, but its convergence stability is insufficient, and it is prone to iterative oscillations. The NSGA-III algorithm has good multi-objective balance, but it has not been optimized for the photovoltaic-noise reduction coupling characteristics, and its optimization efficiency is lower than that of the algorithm in this paper. The MOEA / D algorithm adopts an independent decomposition of multi-objectives, resulting in the worst comprehensive performance.
[0074] Figure 4 This is a graph showing the change in dynamic optimization accuracy during the iteration process of each algorithm in Embodiment 1 of the present invention. Figure 4 As can be seen, the algorithm in this paper achieves an optimization accuracy of over 85% after 40 iterations, surpasses 90% after 80 iterations, and finally converges to an accuracy of 95.3%. The improvement process is smooth without significant oscillations, indicating that the algorithm can quickly approach the global optimum and the optimization process is stable. The conventional NSGA-II algorithm only reaches an accuracy of 85% after 100 iterations, and finally only 82.1%, with significant fluctuations. It exhibits obvious local optimum oscillations between 60 and 80 iterations. The accuracy of the other three algorithms falls between the two. Among them, the NSGA-III algorithm has a higher final accuracy but a slow improvement, reaching 90% accuracy only after 110 iterations. The MOPSO and MOEA / D algorithms improve accuracy faster but have significant fluctuations, with fluctuation amplitudes of 1.02% and 1.35%, respectively. This further proves the performance advantage of the algorithm in this paper.
[0075] III. Overall Performance Comparison The implementation results of this embodiment are compared with those of conventional collaborative optimization schemes, traditional component optimization schemes, and innovative mechanism ablation schemes. The comprehensive performance indicators of each scheme are shown in Table 1. Table 1 Traditional component-specific optimization refers to treating the photovoltaic power generation system and the noise reduction system as two independent subsystems, optimizing them separately. The photovoltaic component independently aims to maximize power generation, while the noise reduction component independently aims to maximize noise reduction. After design, the two components are simply stacked and assembled into a photovoltaic sound barrier without iterative coordination or consideration of their mutual influence. Conventional collaborative optimization places power generation and noise reduction within the same optimization framework for multi-objective collaborative optimization. However, this approach only optimizes the static performance under initial operating conditions and does not consider performance degradation over the entire lifecycle. It is equivalent to optimizing for the same objectives of maximizing power generation efficiency, maximizing noise reduction performance, and minimizing cost, but without incorporating the exponential decay of photovoltaic module output power and the exponential decay of sound absorption coefficients in sound-absorbing materials, and without predicting long-term performance degradation trajectories.
[0076] As shown in Table 1, the proposed solution significantly outperforms other solutions in terms of positive indicators such as net benefit over the entire life cycle and average annual noise reduction, while significantly underperforming other solutions in terms of negative indicators such as carbon emissions over the entire life cycle and static investment payback period. This effectively achieves optimal synergy between economic, environmental, and social benefits. Specifically, the proposed solution achieves a net benefit of RMB 1.286 million over the entire life cycle, representing a 23.8% improvement over conventional synergistic optimization solutions and a 45.2% improvement over traditional component optimization solutions. The average annual noise reduction is 38.2 dB, representing an 11.5% improvement over conventional solutions and a 20.6% improvement over traditional solutions, demonstrating that the synergistic layout of this study effectively improves noise reduction performance. Carbon emissions are reduced by 18.3% compared to conventional solutions, and the investment payback period is shortened by 1.2 years, highlighting the green economics of the proposed solution. Among the comparative solutions, the conventional solution is the second most effective, but due to the lack of attenuation prediction and coupling constraints, its long-term benefits decline significantly, with net benefits decreasing by 8.7% after 18 months compared to the initial value. The innovative mechanism ablation solution's effectiveness falls between the two, further validating the role of the core innovative mechanism. Traditional solutions are the least effective because photovoltaic and noise reduction are designed independently, and their mutual interference leads to poor core functions. The shading of photovoltaic modules reduces noise reduction by 12.3%, and the shading of sound-absorbing and insulating structures reduces power generation efficiency by 9.8%, which fully highlights the necessity of a collaborative optimization strategy.
[0077] Figure 5 This is a comparison chart of the dynamic changes in monthly performance under different operating conditions for various schemes in Embodiment 1 of the present invention. Figure 5It can be seen that under high temperature and high radiation conditions, the monthly power generation of the proposed solution is stable at 175~185 kWh, which is 23.5% higher than that of the conventional solution. Under mild climate conditions, the monthly noise reduction of the proposed solution is stable at 37.5~38.5 dB, which is 11.8% higher than that of the conventional solution, achieving optimal overall efficiency. Under extreme rainstorm conditions, the monthly power generation fluctuation of the proposed solution is only 2.8%, while the fluctuation of the conventional solution is 8.2% and that of the traditional solution is 12.3%, proving that the proposed solution can effectively reduce the impact of rainstorms on power generation and improve the waterproof stability of the system. Under strong wind and sandstorm conditions, the monthly noise reduction attenuation of the proposed solution is only 1.2%, while that of the conventional solution is 4.8% and that of the traditional solution is 6.3%. This is due to the wind and sand resistance design and surface cleaning control of the proposed solution, which reduces the blockage of the sound absorption and insulation structure and the shading of the component surface by wind and sand, ensuring the long-term stable operation of the system.
[0078] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.
Claims
1. A multi-objective optimization method for photovoltaic sound barriers considering performance degradation throughout their entire life cycle, characterized in that: Includes the following steps: S1. Collect basic parameters, including photovoltaic module parameters, sound absorption and insulation structure parameters, support structure parameters, and environmental parameters of the installation site; S2. Based on the basic parameters obtained in S1, the design parameters to be optimized are combined into a decision variable vector, and a multi-objective optimization model including photovoltaic power generation efficiency, noise reduction performance and life cycle cost is constructed, and the constraints of the multi-objective optimization model are set. S3. Introduce the time dimension and construct a full life cycle attenuation prediction model that includes photovoltaic module output power attenuation, sound absorption material sound absorption coefficient attenuation and coupling attenuation factor attenuation. Embed the full life cycle attenuation prediction model into the multi-objective optimization model constructed in S2 to obtain a multi-objective optimization model with embedded performance attenuation characteristics. S4. The improved IA-NSGA-Ⅱ algorithm, which includes a penalty mechanism and an adaptive collaborative convergence strategy, is used to solve the multi-objective optimization model of the S3 embedded full life cycle decay prediction model, and the Pareto optimal solution set that takes into account the synergistic optimization of power generation, noise reduction and long-term economic efficiency is obtained. S5. Select the optimal design scheme from the Pareto optimal solution set output by S4, and carry out the structural design, component and construction installation of the photovoltaic sound barrier according to the design scheme; S6. After the photovoltaic sound barrier is put into actual operation, collect on-site operation data regularly, compare it with the output of the full life cycle attenuation prediction model, correct the model parameters online according to the deviation, and feed the corrected parameters back to the attenuation prediction model for updating subsequent predictions.
2. The multi-objective optimization method for photovoltaic sound barriers considering performance degradation throughout their entire life cycle, as described in claim 1, is characterized in that: The basic parameters in S1 include: Photovoltaic module parameters include photovoltaic module conversion efficiency, photovoltaic module weight per unit area, photovoltaic module tilt angle, photovoltaic module installation area, photovoltaic module cleaning cycle, and tilt angle adjustment frequency; The structural parameters for sound absorption and insulation include the thickness of the sound-absorbing layer, the thickness of the sound insulation layer, the perforation rate of the sound-absorbing material, the sound absorption coefficient of the sound-absorbing material, the density of the sound insulation material, and the weight per unit volume of the sound-absorbing material. Support structure parameters include the maximum load-bearing capacity of the support structure; Environmental parameters for the installation site include annual average solar irradiance, air characteristic impedance, and the speed of sound in the air.
3. The multi-objective optimization method for photovoltaic sound barriers considering performance degradation throughout their entire life cycle, as described in claim 1, is characterized in that: The specific process of S2 is as follows: S2.1 Determine the decision variable vector X The design parameters to be optimized are combined into an 8-dimensional decision variable vector. X , is represented as: In the formula, α Photovoltaic module tilt angle; S This refers to the actual area where the photovoltaic modules are installed. η p For component conversion efficiency; d 1 represents the thickness of the sound-absorbing layer; d 2 represents the thickness of the sound insulation layer; The perforation rate of the sound-absorbing material; T For the cleaning cycle of photovoltaic modules; f Adjust the frequency to adjust the tilt angle; S2.2 Constructing a multi-objective optimization function The constructed photovoltaic sound barrier power generation-noise reduction strongly coupled multi-objective optimization model is as follows: In the formula, E year The annual power generation of the photovoltaic system is used as a quantitative indicator. H avg Annual average irradiance; γ is the overall efficiency coefficient of the photovoltaic system; δ t This is the temperature decay coefficient; Y For cumulative years; The annual degradation rate of photovoltaic modules; This indicates the amount of electricity generated in the first year; The equivalent number of years for decay; N avg The annual noise reduction amount is a quantitative indicator. N abs This refers to the noise reduction amount of the sound-absorbing layer; N ins This represents the noise reduction amount of the sound insulation layer; Δ N Noise reduction loss due to shading of photovoltaic modules; This is the attenuation term of the sound-absorbing layer performance index. These are the preset equivalent acoustic parameters or static correction coefficients related to the geometric configuration of the sound-absorbing structure; For the occlusion loss growth term; γ d The annual growth rate of the loss; t For time; C init For initial investment cost, C op For operation and maintenance costs, C en Energy consumption cost; C LCC Cost over the entire lifecycle; Set constraints The boundary constraints for the decision variables are: The performance constraints are: Structural safety constraints are: In the formula, ρ p This refers to the weight per unit area of the photovoltaic module. S total This represents the total usable surface area of the sound barrier. ρ Density of sound insulation material; ρ abs The weight per unit volume of the sound-absorbing material; W max To support the maximum load-bearing capacity of the structure; Coupling and Coordination Constraints To quantify and suppress the power generation-noise reduction coupling conflict, a coupling conflict coefficient is introduced. The calculation formula is: In the formula, P ( α () represents the actual power generation. N ( α The tilt angle of the photovoltaic module is... α The actual noise reduction amount at that time; P 0 represents the ideal power generation capacity of a photovoltaic module; N 0 represents the ideal noise reduction level for the noise reduction system; Coupling conflict coefficient The constraints that need to be satisfied are: .
4. The multi-objective optimization method for photovoltaic sound barriers considering performance degradation throughout their entire life cycle, as described in claim 1, is characterized in that: The full lifecycle degradation prediction model constructed by S3 is as follows: The output power of photovoltaic modules decreases exponentially over time, as shown in the formula: in, P ( t ) No. t Actual annual output power; The attenuation of the sound absorption coefficient of sound-absorbing materials is affected by dust blockage and moisture erosion, as shown in the formula: In the formula, For the first t The actual sound absorption coefficient for the year; The initial sound absorption coefficient; β s The annual attenuation rate of the sound absorption coefficient of the sound-absorbing material. β s Preferred rate: 0.3% / year; α res The residual sound absorption coefficient; The two attenuation methods are coupled through the attenuation factor. ξ ( t Related: In the formula, This is the coupling attenuation coefficient; The full life cycle degradation prediction model is embedded into the multi-objective optimization model constructed by S2. The coupling degradation factor associates the independent degradation processes of photovoltaics and noise reduction into a mutual feedback degradation process. The original static objective is replaced by the cumulative power generation efficiency, cumulative noise reduction performance and full life cycle cost as optimization objectives.
5. The multi-objective optimization method for photovoltaic sound barriers considering performance degradation throughout their entire life cycle, as described in claim 1, is characterized in that: The process of solving the multi-objective optimization model of the S3 embedded full life cycle decay prediction model using the improved IA-NSGA-Ⅱ algorithm in S4 is as follows: S4.1 Real Number Encoding: Each individual is encoded as an 8-dimensional real number vector, which is linked to the decision variable vector. X =[ α , S , η p , d 1, d 2, , T , f ] T There is a one-to-one correspondence, and the range of values for each dimension is consistent with the boundary constraints of the decision variables; S4.2 Population initialization: Randomly generate several individuals in the feature space, each individual representing a photovoltaic sound barrier design scheme; S4.3 Introduce a coupling penalty mechanism: Set the coupling penalty factor ω to: In the formula, k p This is the penalty coefficient; λ The coupling conflict coefficient; The coupling penalty factor is incorporated into the following three objective functions for modification, resulting in the following fitness function: (1) Power generation efficiency fitness function for: (2) Noise reduction performance fitness function for: (3) Cost fitness function for: In the formula, F 1. F 2. F 3 represents the corrected fitness function value; C LCC,max This represents the maximum value of the total lifecycle cost. S4.4 Design an adaptive collaborative convergence strategy Design an adaptive collaborative convergence strategy to dynamically adjust the crossover probability. With the probability of mutation It can quickly search for the global optimum in the early stages of iteration and converge precisely in the later stages of iteration. (1) Adaptive crossover probability The dynamic adjustment formula is as follows: In the formula, p c For adaptive crossover probability; p c0 The initial crossover probability; k c This is the crossover probability adjustment coefficient; n This represents the current iteration step number; G max The maximum number of iterations; (2) Adaptive mutation probability The dynamic adjustment formula is as follows: The symbols in the formula are defined as follows: For adaptive mutation probability; p m0 The initial mutation probability; k m This is the adjustment factor for the probability of mutation; S4.5 Population Evolution: Based on the corrected fitness function values, perform non-dominated sorting and crowding calculations on the individuals in the population, and sequentially execute binary tournament selection to choose parent individuals; when the generated random number for each pair of parents is less than or equal to the adaptive crossover probability... Then perform simulated binary crossover; otherwise, directly copy. When the generated random number for each dimension of each offspring is less than or equal to the adaptive mutation probability... Then, polynomial mutation is performed to generate a new generation of population; S4.6 Repeat the iteration until the maximum number of iterations G is reached. max Output the Pareto optimal solution set.
6. The multi-objective optimization method for photovoltaic sound barriers considering life-cycle performance degradation as described in claim 5, characterized in that: The online correction model parameters in S6 include the photovoltaic module degradation rate. β p Attenuation rate of sound-absorbing materials β s and coupling attenuation coefficient The specific correction method is as follows: For photovoltaic module degradation rate β p Using periodically collected actual power generation data, based on the formula The least squares method or exponential regression is used for fitting, and the results are dynamically updated. β p The measured estimated value; Annual attenuation rate of sound absorption coefficient of sound-absorbing materials β s Based on periodic measured data of sound absorption coefficient or noise reduction behind the barrier, and using the formula... Perform exponential fitting, after subtracting the residual sound absorption coefficient α res Re-identification β s ; For coupling attenuation coefficient Based on coupling attenuation factor By using periodically collected synchronous observation data of power generation and noise reduction, As the dependent variable, Perform linear regression on the independent variable and re-estimate the regression coefficients. value.
