Park carbon emission optimization method based on BIM and online genetic algorithm

Abstract

The invention relates to a park carbon emission optimization method based on BIM and an online genetic algorithm, and the method comprises the following steps: S1, obtaining static data through a BIM model, and carrying out the data cleaning, standardization and alignment of the input static data; s2, acquiring internal dynamic data and external environment dynamic data; s3, unifying timestamps and formats through a digital twin platform, eliminating noise, and carrying out standardization and alignment processing on input data; s4, designing a multi-objective optimization model according to the park engineering background; s5, initializing a population of a genetic algorithm, and encoding chromosomes; s6, calculating indexes of carbon emission, cost and comfort of each individual, and calculating a fitness function; s7, performing selection, crossover and variation by adopting a roulette according to the fitness function; and S8, selecting an individual with the highest fitness from the current population as an optimization strategy. According to the method, the carbon emission reduction efficiency in a complex environment is remarkably improved, and the method is suitable for sustainable management of the whole life cycle of a building.

Description

Technical Field

[0001] This invention relates to the field of intelligent building and environmental monitoring technology, and more specifically, to a method for optimizing carbon emissions in industrial parks based on BIM and online genetic algorithms. Background Technology

[0002] The increasing complexity of construction projects often necessitates the integration and multidimensionality of building models, making Building Information Modeling (BIM) an indispensable part of the construction industry. BIM technology can optimize resource use and simulate energy consumption, while also serving as a data foundation for simulating the selection of building materials and construction techniques to reduce carbon emissions during construction. However, currently, BIM models are primarily used for static design and analysis throughout the building's entire lifecycle, lacking deep integration with real-time dynamic data (such as energy consumption and carbon emission monitoring data collected by sensors). Specifically, optimization strategies based on static models cannot respond in real-time to environmental changes (such as fluctuations in air conditioning load caused by sudden weather changes), and carbon emission prediction relies on offline simulations, resulting in discrepancies with actual operating conditions.

[0003] In addition, traditional genetic algorithms (GA) have always suffered from insufficient dynamic adaptability in optimizing carbon emission scheduling strategies. As a result, traditional genetic algorithms rely on fixed parameters (such as crossover probability and mutation rate) and offline iteration, making it difficult to respond to environmental changes (such as carbon price fluctuations and equipment failures) in real time. Therefore, when applied to dynamic scenarios, they are prone to getting trapped in local optima, resulting in suboptimal carbon emission optimization results. Moreover, the computational latency is high, which cannot meet the requirements of real-time control.

[0004] It is worth noting that although traditional genetic algorithms have global search capabilities, they suffer from convergence difficulties in high-dimensional parameter spaces. Furthermore, monitoring and optimizing carbon emissions at the park level involves thousands of variables (such as building and equipment parameters, traffic routes, and energy dispatch strategies), causing the computational complexity of traditional GA to increase exponentially, resulting in long iteration times and failing to meet real-time requirements. In addition to addressing the convergence problem, the ability to handle constraints must also be considered, i.e., dynamically updating constraints such as carbon emission limits and equipment capacity. Summary of the Invention

[0005] The technical problem to be solved by this invention is to provide a method for optimizing carbon emissions in industrial parks based on BIM and online genetic algorithms. This method can combine real-time data, dynamic parameter adjustment, multi-objective optimization and adaptive mechanisms to achieve accurate and efficient optimization of carbon emissions in industrial parks.

[0006] The technical solution adopted by this invention to solve its technical problem is: constructing a method for optimizing carbon emissions in industrial parks based on BIM and online genetic algorithms, including the following steps: S1. Obtain static data through the BIM model, and then perform data cleaning, standardization and alignment on the input static data; S2. Obtain internal dynamic data and external environment dynamic data; S3. By unifying timestamps and formats through a digital twin platform, noise is eliminated, input data is standardized and aligned, key carbon emission-related parameters are extracted, and non-core data is simplified. S4. Construct a multi-objective dynamic optimization model that supports real-time weight adjustment, and combine the mass balance method and the measurement method to calculate carbon emissions and dynamically adjust weights.

[0007] S5. Initialize the population of the genetic algorithm and encode the chromosomes, where each individual represents an optimization strategy, including equipment operating parameters, energy scheduling strategies, and traffic route optimization parameters; S6. Combine BIM static data with real-time dynamic data to calculate the carbon emissions, costs, and comfort indicators for each individual, and calculate the fitness function. S7. Based on the fitness function, roulette wheel selection, crossover, and mutation are performed, with priority given to retaining individuals with high fitness. S8. Based on environmental perception and constraint adaptation, dynamically adjust parameters and select the individual with the highest fitness from the current population as the optimization strategy.

[0008] According to the above scheme, in step S1, the static data includes building structure, equipment parameters, material carbon footprint, and industrial production process parameters.

[0009] According to the above scheme, in step S2, the internal dynamic data includes building energy consumption, temperature and humidity, traffic flow, and equipment status; the external environmental dynamic data includes weather forecasts, carbon price fluctuations, and policy and regulatory updates.

[0010] According to the above scheme, in step S3, key carbon emission-related parameters are extracted, including equipment operating status and energy consumption. According to the above scheme, in step S4, the multi-objective dynamic optimization model is as follows:

[0011] st Carbon Cap (Dynamic carbon emission cap) Demand (Energy supply and demand balance) in, , , This is a weighting factor for the three types of carbon emission outputs; This represents total carbon emissions. Represents the total cost. Indicating a loss of comfort, Carbon Cap represents the system's actual carbon emissions in time period t, obtained by summing the products of the actual power generation of all energy technologies during that time period and their carbon emission coefficients; The dynamic carbon emission cap for time period t is an exogenous parameter that changes over time, reflecting either a gradual tightening of policies or seasonal differences. Demand represents the total energy supply of the system in time period t, in MWh, obtained by summing the actual power generation of all energy technologies during that time period. This represents the total energy demand for time period t, in MWh, given by load forecasting, and is an exogenous parameter that varies over time.

[0012] According to the above scheme, in step S4, carbon emissions are calculated by combining the mass balance method and the actual measurement method: Mass balance method:

[0013] Actual measurement method: CEMS measured concentration Flow rate time High carbon emission weight: The weights are dynamically adjusted as follows: Energy saving priority: w1=0.75, w2=0.25 Emissions priority: w1=0.25, w2=0.75; Overall optimization: w1=w2=0.5.

[0014] According to the above scheme, in step S6, the fitness function is:

[0015] in, This is the dynamic penalty coefficient.

[0016] According to the above scheme, in S7, the dynamic crossover probability is set as follows:

[0017] in, Here, represents the initial dynamic crossover probability, and diversity is a population diversity index. These are the weighting coefficients.

[0018] According to the above scheme, in S7, the dynamic variation rate is:

[0019] in, For the number of iterations, To adjust the parameters, the mutation rate is reduced as the number of iterations increases in order to accelerate convergence.

[0020] The present invention also provides an electronic device, comprising: a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; The memory stores a computer program that, when executed by the processor, causes the processor to perform the steps of the park carbon emission optimization method based on BIM and online genetic algorithm.

[0021] The BIM-based and online genetic algorithm-based method for optimizing carbon emissions in industrial parks, as described in this invention, has the following beneficial effects: 1. This invention integrates the static data of BIM with the dynamic optimization capabilities of genetic algorithms, and combines real-time data feedback and adaptive parameter adjustment strategies to achieve real-time monitoring, multi-objective optimization and closed-loop control of carbon emissions in the park, which significantly improves carbon emission reduction efficiency in complex environments and is suitable for sustainable management of the entire building life cycle. 2. This invention enables dynamic optimization driven by real-time data, such as real-time response to environmental factors like carbon price fluctuations and weather changes, avoiding the lag of traditional offline algorithms. Simultaneously, through dynamic weight adjustments, it flexibly balances emission reduction, cost, and comfort, minimizing carbon emissions in the industrial park while balancing multiple objectives including cost and user comfort. For high-precision carbon emission accounting, it combines the industrial process mass balance method and combustion emission measurement method to ensure the accuracy of emission calculations. Attached Figure Description

[0022] The present invention will be further described below with reference to the accompanying drawings and embodiments. In the accompanying drawings: Figure 1 This is a flowchart of the carbon emission optimization method for industrial parks based on BIM and online genetic algorithms, as described in this invention. Detailed Implementation

[0023] To provide a clearer understanding of the technical features, objectives, and effects of the present invention, specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0024] Example 1 like Figure 1 As shown, the carbon emission optimization method for industrial parks based on BIM and online genetic algorithms of the present invention includes the following steps: S1. Data Acquisition and Preprocessing: Static data is acquired through the BIM model. The static data includes building structure, equipment parameters, material carbon footprint, and industrial production process parameters. The material carbon footprint includes air conditioning energy efficiency ratio and photovoltaic power generation efficiency. The input static data is then cleaned, standardized, and aligned.

[0025] S2. Based on IoT sensor data, obtain dynamic data such as building energy consumption, temperature and humidity, traffic flow, equipment status, as well as external environmental dynamic data such as weather forecasts, carbon price fluctuations, and policy and regulatory updates, with the policies and regulations referring to carbon emission quotas.

[0026] S3. A digital twin platform is used to standardize timestamps and formats, eliminate noise, standardize and align input data, extract key carbon emission-related parameters, and simplify non-core data. For outlier detection, a linear regression model is used to fill in missing values, and key carbon emission-related parameters, including equipment operating status and energy consumption, are extracted. S4. Based on the background of the park project, construct a multi-objective dynamic optimization model that supports real-time weight adjustment, and combine the mass balance method and the actual measurement method to calculate carbon emissions and dynamically adjust weights.

[0027] The multi-objective dynamic optimization model is as follows:

[0028] st Carbon Cap (Dynamic carbon emission cap) Demand (Energy supply and demand balance) in, , , This is the weighting coefficient for the three types of carbon emission outputs. This represents total carbon emissions. Represents the total cost. Indicating a loss of comfort, Carbon Cap represents the system's actual carbon emissions during time period t, obtained by summing the products of the actual power generation of all energy technologies during that time period and their carbon emission coefficients. This represents the dynamic carbon emission cap for time period t, an exogenous parameter that changes over time and reflects either a gradual tightening of policies or seasonal differences. This represents the total energy supply of the system in time period t (unit: MWh), obtained by summing the actual power generation of all energy technologies during that time period. The total energy demand for time period t (in MWh) is an exogenous parameter given by load forecast and varying over time.

[0029] Carbon emissions are calculated by combining the mass balance method and the measured method: (Mass balance method) CEMS measured concentration Flow rate Time (measured method) The weights are dynamically adjusted as follows: Energy conservation priority: w1=0.75, w2=0.25 (higher carbon emission weight). Emissions priority: w1=0.25, w2=0.75.

[0030] Overall optimization: w1=w2=0.5.

[0031] S5. Initialize the genetic algorithm population by encoding chromosomes, where each individual represents an optimization strategy, including: equipment operating parameters, energy scheduling strategy, and traffic route optimization parameters. Set the initial population size to N=100.

[0032] S6. Fitness Assessment. By combining BIM static data with real-time dynamic data, calculate the carbon emissions, cost, comfort, and other indicators for each individual, and calculate the fitness function.

[0033] The fitness function is:

[0034] in, This is a dynamic penalty coefficient, which increases when carbon prices rise.

[0035] S7. Selection, crossover, and mutation are performed based on the fitness function. The selection mechanism adopts roulette wheel selection, which prioritizes the retention of individuals with high fitness.

[0036] The dynamic crossover probability is set as follows:

[0037] in, Here, represents the initial dynamic crossover probability, and diversity is a measure of population diversity, such as the degree of difference in parameters among individuals. These are the weighting coefficients.

[0038] Dynamic variation rate:

[0039] in, For the number of iterations, To adjust the parameters. As the number of iterations increases, the mutation rate is reduced to accelerate convergence.

[0040] In dynamic parameter adjustment, when an increase in carbon price is detected, the parameter is automatically reduced. To maintain population diversity, priority should be given to exploring low-carbon solutions, while carbon emission caps should be updated in real time. And energy supply and demand balance constraints (such as switching to backup energy when the power grid fails).

[0041] S8. Dynamically adjust parameters based on environmental perception and constraint adaptation. Select the individual with the highest fitness from the current population as the optimization strategy.

[0042] Taking the actual operation of a smart industrial park in East China from 14:00 to 15:00 on July 15, 2025 as an example: the park's BIM model has integrated parameters of 3 office buildings, 1 data center, and a photovoltaic-energy storage system; the IoT platform collected real-time data showing a total load of 5.2 MW, photovoltaic output of 1.8 MW, and an average indoor temperature of 26.5℃. On that day, the local carbon quota was tightened by 10%, with an hourly carbon emission limit of 1.8 tons of CO2.

[0043] Based on the method of this invention, the digital twin platform constructs a multi-objective optimization model, setting weights. (Carbon emissions) (Economic costs) (Comfort level). An online genetic algorithm (population size 100, maximum iterations 50 generations, dynamic crossover / mutation rate) is used to solve the problem, and the final scheduling strategy is as follows: 1. The air conditioner setting temperature has been lowered from 27℃ to 25.5℃; 2. The energy storage system has a discharge power of 1.0 MW (it had already been charged to full capacity at noon). 3. The number of electric shuttle buses operating in the park has been reduced from 20 to 17; 4. Keep the diesel generator off.

[0044] After implementation, the actual carbon emissions during that period were 1.72 tons of CO2, lower than the upper limit; the electricity cost was 2,830 yuan, only 4.1% higher than the baseline strategy; the indoor temperature remained stable at 25.8℃, meeting user comfort standards. The overall carbon emissions for that day decreased by 12.3% compared to the unoptimized scenario, verifying the effectiveness and practicality of the invention in a real engineering environment.

[0045] I. Carbon Emission Calculation The park's total carbon emissions are contributed by electricity purchased from the grid and backup diesel generators (if activated), while photovoltaic and energy storage are zero-carbon.

[0046] Let the electricity purchased by the power grid during time period t be: Diesel power generation: Carbon emission factors of power grids: Diesel carbon emission factor: The carbon emissions for that hour are: .

[0047] The carbon emissions for that hour are:

[0048] The optimized strategy is as follows: Power grid purchases will be reduced to...

[0049] Optimized carbon emissions:

[0050] II. Economic Cost Calculation Assuming the electricity price structure is time-of-use pricing (14:00-15:00 is peak time): Electricity price:

[0051] Energy storage charging and discharging efficiency: However, during this period, only discharge occurs, and the cost has already been factored into the cost of previous charging (midday off-peak electricity). ) Cost of diesel generator:

[0052] Cost before optimization:

[0053] After optimization (power purchased from the grid: 2400 kW):

[0054] The "opportunity cost" of energy storage discharge is factored in: if the electricity is discharged during off-peak hours... After filling, the equivalent savings are:

[0055] However, it should be noted that energy storage and discharge... The "opportunity cost" should be included. Full lifecycle cost accounting should be adopted, or service fees and maintenance surcharges should be included. Actual measured value.

[0056] III. Calculation of Comfort Penalty Items Define an inappropriate penalty function (linear overrun penalty):

[0057] in This is the penalty coefficient.

[0058] Before optimization:

[0059] After optimization:

[0060] IV. Calculation of Objective Function Value Using a weighted normalization approach, the units of each indicator are first standardized and normalized (with reference to the benchmark strategy): Set a baseline strategy (unoptimized):

[0061]

[0062]

[0063] The optimized relative index is as follows:

[0064]

[0065]

[0066] Objective function (the smaller the better):

[0067] V. Fitness Function of Genetic Algorithm Introduce penalties for violating constraints. Set a carbon cap. ,current No violation.

[0068] Fitness is defined as:

[0069] in

[0070] Therefore:

[0071] This individual has high fitness in the population and is easily selected for retention.

[0072] VI. Verification of Emission Reduction Effect Daily emission reduction rate calculation (based on case data):

[0073] If the base date emissions are ,but:

[0074] Example 2 The present invention also provides an electronic device, comprising: a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; The memory stores a computer program that, when executed by the processor, causes the processor to perform the steps of the park carbon emission optimization method based on BIM and online genetic algorithm.

[0075] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. A method for optimizing carbon emissions in industrial parks based on BIM and online genetic algorithms, characterized in that, Includes the following steps: S1. Obtain static data through the BIM model, and then perform data cleaning, standardization and alignment on the input static data; S2. Obtain internal dynamic data and external environment dynamic data; S3. By unifying timestamps and formats through a digital twin platform, noise is eliminated, input data is standardized and aligned, key carbon emission-related parameters are extracted, and non-core data is simplified. S4. Construct a multi-objective dynamic optimization model that supports real-time weight adjustment, and combine the mass balance method and the measurement method to calculate carbon emissions and dynamically adjust weights. S5. Initialize the population of the genetic algorithm and encode the chromosomes, where each individual represents an optimization strategy, including equipment operating parameters, energy scheduling strategies, and traffic route optimization parameters; S6. Combine BIM static data with real-time dynamic data to calculate the carbon emissions, costs, and comfort indicators for each individual, and calculate the fitness function. S7. Based on the fitness function, roulette wheel selection, crossover, and mutation are performed, with priority given to retaining individuals with high fitness. S8. Based on environmental perception and constraint adaptation, dynamically adjust parameters and select the individual with the highest fitness from the current population as the optimization strategy.

2. The method for optimizing carbon emissions in industrial parks based on BIM and online genetic algorithms according to claim 1, characterized in that, In step S1, the static data includes building structure, equipment parameters, material carbon footprint, and industrial production process parameters.

3. The method for optimizing carbon emissions in industrial parks based on BIM and online genetic algorithms according to claim 1, characterized in that, In step S2, the internal dynamic data includes building energy consumption, temperature and humidity, traffic flow, and equipment status; the external environmental dynamic data includes weather forecasts, carbon price fluctuations, and policy and regulatory updates.

4. The method for optimizing carbon emissions in industrial parks based on BIM and online genetic algorithms according to claim 1, characterized in that, In step S3, key carbon emission-related parameters are extracted, including equipment operating status and energy consumption.

5. The method for optimizing carbon emissions in industrial parks based on BIM and online genetic algorithms according to claim 1, characterized in that, In step S4, the multi-objective dynamic optimization model is as follows: st Carbon Cap (Dynamic carbon emission cap) Demand (Energy supply and demand balance) in, , , This is a weighting factor for the three types of carbon emission outputs; This represents total carbon emissions. Represents the total cost. Indicating a loss of comfort, Carbon Cap represents the system's actual carbon emissions in time period t, obtained by summing the products of the actual power generation of all energy technologies during that time period and their carbon emission coefficients; The dynamic carbon emission cap for time period t is an exogenous parameter that changes over time, reflecting either a gradual tightening of policies or seasonal differences. Demand represents the total energy supply of the system in time period t, in MWh, obtained by summing the actual power generation of all energy technologies during that time period. This represents the total energy demand for time period t, in MWh, given by load forecasting, and is an exogenous parameter that varies over time.

6. The method for optimizing carbon emissions in industrial parks based on BIM and online genetic algorithms according to claim 5, characterized in that, In step S4, carbon emissions are calculated by combining the mass balance method and the measured method: Mass balance method: Actual measurement method: CEMS measured concentration Flow rate time High carbon emission weight: The weights are dynamically adjusted as follows: Energy saving priority: w1=0.75, w2=0.25 Emissions priority: w1=0.25, w2=0.75; Overall optimization: w1=w2=0.

5.

7. The method for optimizing carbon emissions in industrial parks based on BIM and online genetic algorithms according to claim 1, characterized in that, In S6, the fitness function is: in, This is the dynamic penalty coefficient.

8. The method for optimizing carbon emissions in industrial parks based on BIM and online genetic algorithms according to claim 7, characterized in that, In S7, the dynamic crossover probability is set as follows: in, Here, represents the initial dynamic crossover probability, and diversity is a population diversity index. These are the weighting coefficients.

9. The method for optimizing carbon emissions in industrial parks based on BIM and online genetic algorithms according to claim 8, characterized in that, In S7, the dynamic mutation rate is: in, For the number of iterations, To adjust the parameters, the mutation rate is reduced as the number of iterations increases in order to accelerate convergence.

10. An electronic device, characterized in that, include: The processor, communication interface, memory, and communication bus are connected, with the processor, communication interface, and memory communicating with each other via the communication bus. The memory stores a computer program that, when executed by the processor, causes the processor to perform the steps of the method according to any one of claims 1 to 9.

Images