Park carbon emission influence factor scene construction method and system

Abstract

The invention discloses a park carbon emission influence factor scene construction method and system. The method comprises the steps of collecting park data, and constructing a carbon emission influence factor scene based on a self-optimization k-means + + clustering method. And performing modeling calculation on each influence factor. And calculating the total carbon emission, and optimizing the clustering result. According to the method, through clustering analysis, the carbon emission influence factors of different functional spaces can be classified and subjected to refined analysis. And the Markov chain model is introduced, so that the in-room states and on-off actions of personnel in different functional spaces can be accurately simulated, and the energy consumption change can be more accurately predicted and managed. Energy consumption fluctuation in the park can be effectively smoothed by storing energy in the low ebb period and releasing energy in the peak period. Besides, the intelligent control system can dynamically adjust operation of the equipment, and energy consumption and carbon emission are reduced to the maximum extent while the comfort level is met in combination with personnel behavior prediction and real-time data.

Description

Technical Field

[0001] The present invention relates to the technical field of park scene construction, and specifically to a method and system for constructing a scene of factors affecting carbon emissions in a park. Background Art

[0002] Low-carbon parks use smart energy management technology and multi-energy systems to effectively control and optimize carbon emissions. In these parks, data-driven carbon footprint tracking and analysis has become a core technical means, which can promote the realization of the park's low-carbon goals by quantifying carbon emission factors and optimizing energy management solutions.

[0003] Data-driven carbon footprint tracking and analysis technology plays a key role in low-carbon parks. By real-time monitoring and recording of energy consumption and emission data, dynamic tracking and analysis of carbon emissions can be achieved. This method can not only quantify carbon emission factors, but also promote the realization of low-carbon goals in the park by optimizing energy management solutions. In summary, the development of low-carbon parks not only depends on policy support and technological innovation, but also requires scientific carbon emission monitoring and management methods. Through intelligent energy management technology and data-driven carbon footprint tracking and analysis, low-carbon parks can achieve effective control and optimization of carbon emissions and promote the realization of a modern sustainable development model.

[0004] Although data-driven carbon footprint tracking and analysis technology plays an important role in low-carbon parks and sustainable development, there are still some existing problems and challenges:

[0005] There are few studies, mainly focusing on energy consumption: So far, there are relatively few studies on carbon emission monitoring and carbon footprint tracking, and most of the studies focus on energy consumption. This has led to the fact that the technologies and methods for carbon emission monitoring and carbon footprint tracking are not mature and comprehensive enough.

[0006] Park-level monitoring and tracking are difficult: Due to the large number of systems at the park level, monitoring and tracking are difficult. At this stage, research and application are more focused on single buildings, while for complex park-level systems, monitoring and tracking technologies and methods need to be further developed and improved.

[0007] Lack of systematic scientific methods: At present, there is a lack of systematic scientific methods for carbon emission monitoring and carbon footprint tracking. Although some methods and technologies are being applied, there is no unified standard and system, which leads to certain difficulties and challenges in practical application.

[0008] Lack of detailed analysis of specific influencing factors: Current carbon emission monitoring and carbon footprint tracking research lacks detailed analysis of specific influencing factors. Most studies are usually based on overall energy consumption or large-scale energy consumption patterns, while ignoring more refined influencing factors, such as personnel behavior, equipment operating status, weather changes, etc. There are significant differences in the impact of different building functional spaces, electrical equipment, time periods, weather conditions, etc. on carbon emissions, but these differences are often not fully considered and modeled in existing studies. This rough analysis method may lead to insufficient accuracy in carbon emission prediction and monitoring, and fail to provide targeted carbon emission reduction recommendations.

[0009] These issues indicate that although data-driven carbon footprint tracking and analysis technologies have important potential in low-carbon parks, further research and technological innovation are still needed to address existing challenges. Summary of the invention

[0010] In view of the above-mentioned problems, the present invention is proposed.

[0011] Therefore, the technical problems solved by the present invention are: There are relatively few studies, mainly focusing on energy consumption: Up to now, there are relatively few studies on carbon emission monitoring and carbon footprint tracking, and most of the studies are mainly focused on energy consumption. This has led to the fact that the technologies and methods for carbon emission monitoring and carbon footprint tracking are not mature and comprehensive enough. The current carbon emission monitoring and carbon footprint tracking research lacks detailed analysis of specific influencing factors. Most studies are usually based on overall energy consumption or large categories of energy consumption patterns, while ignoring more refined influencing factors, such as personnel behavior, equipment operating status, weather changes, etc.

[0012] To solve the above technical problems, the present invention provides the following technical solutions: a method for constructing scenarios of factors affecting carbon emissions in a park, comprising: collecting park data and constructing scenarios of factors affecting carbon emissions based on a self-optimizing k-means++ clustering method.

[0013] Model and calculate each influencing factor.

[0014] Calculate total carbon emissions and optimize clustering results.

[0015] As a preferred solution of the method for constructing scenarios of factors affecting industrial park carbon emissions described in the present invention, the self-optimizing k-means++ clustering method includes automatically finding the optimal K value in combination with SC and DBI indicators, and using the k-means++ method to optimize the initial clustering center. The steps include finding the optimal K value, initializing the clustering center, classifying samples, and updating the clustering center.

[0016] Repeat sample classification and cluster center update until the cluster center no longer changes.

[0017] As a preferred solution of the method for constructing the scenario of the factors affecting the carbon emissions of the park described in the present invention, the initial cluster center includes randomly selecting a sample as the first cluster center during the cluster center initialization process, and calculating the shortest distance between each sample and the current existing center. The next cluster center is selected according to the probability of each sample being selected as the next center, until the number of selected cluster centers reaches K, and finally the cluster center initialization is terminated.

[0018] In the clustering process, the optimal K value is first determined, and clustering is started according to the steps of cluster center initialization. Each sample is assigned to the cluster closest to it, the average value of the data in each cluster is calculated, and the cluster center is updated.

[0019] Check whether the cluster center has changed. If so, repeat the cluster center initialization and clustering process. Otherwise, output the clustering result.

[0020] As a preferred scheme of the method for constructing scenarios of factors affecting park carbon emissions described in the present invention, the method for constructing scenarios of factors affecting park carbon emissions based on the self-optimizing k-means++ clustering method includes applying the self-optimizing K-means++ algorithm to the scenario construction of factors affecting park carbon emissions, and performing cluster analysis on the factors affecting park carbon emissions through the algorithm.

[0021] When initializing the cluster center, a carbon emission influencing factor is randomly selected as the initial cluster center, the distance between other influencing factors and the center is calculated, and other cluster centers are selected according to the probability of the distance to ensure that the selected center can represent a variety of influencing factors.

[0022] In the clustering process, various influencing factors are grouped and clustered, including: indoor factors, time factors, weather factors and calendar factors.

[0023] The accuracy of clustering results is ensured by calculating the average carbon emission contribution value in each cluster, updating the cluster center, and checking whether the center has changed.

[0024] As a preferred solution of the method for constructing scenarios of factors affecting carbon emissions in the park described in the present invention, the modeling and calculation of each influencing factor includes: indoor factors use Markov chains to describe the presence of people in the room. For a single-person office, the presence of people in the room is simulated by sampling a Markov chain. For a multi-person room, multiple Markov chains are superimposed to simulate the overall occupancy of the room. Assuming that the step length staying in state i is ST i , then ST i It obeys the geometric distribution, expressed as:

[0025] P{ST i =k}=p ik-1 (1-p i )

[0026] Among them, P{ST i =k} ​​represents the probability that an individual starts in state i and leaves state i after k steps and i steps.

[0027] The park is divided into multiple functional spaces, each of which has different sensitivities to different weather factors. According to the weights of each subsystem and influencing factors, a corresponding scenario model is constructed for each functional space:

[0028] Based on the collected key weather data and building energy consumption data, the relationship between weather factors and energy consumption is established. Different functional spaces in the park have different sensitivities to weather factors. Through weight analysis, sensitive factors are identified and a prediction model is constructed.

[0029] Establishing the correlation between weather factors and energy consumption:

[0030]

[0031] Among them, E i (t) represents the energy consumption of functional space i at time t, F j (t) represents the value of weather factor j at time t, w ij is the weight of function space i on weather factor j.

[0032] In order to achieve dynamic optimization of energy-consuming equipment, a real-time feedback and prediction mechanism is established, combined with a machine learning prediction model, so that the equipment can be adjusted in advance under different weather scenarios to avoid unnecessary energy consumption fluctuations, expressed as:

[0033] P(t+1)=α·P(t)+β·E pred (t+1)

[0034] Where P(t) represents the power demand of the device at time t, E pred (t+1) represents the energy consumption forecast for the future time point t+1 based on the weather forecast model, α represents the adjustment coefficient, and β represents the adjustment coefficient, which is used to balance the weight of historical data and predicted data.

[0035] For each functional space, calculate its carbon emissions:

[0036]

[0037] Among them, C i (t) represents the carbon emissions of functional space i at time t, η k represents the carbon emission coefficient of k energy subsystems, E ik (t) represents the energy consumption of the kth subsystem used by functional space i at time t.

[0038] In order to smooth the energy consumption of each functional space and avoid a sudden increase in carbon emissions, the coordinated scheduling of the energy storage system is introduced:

[0039] S(t)=S(t-1)+P charge (t)-P discharge (t)

[0040] Where S(t) represents the energy storage level of the energy storage system at time t, P charge (t) represents the charging power of the energy storage system, P discharge (t) represents the discharge power of the energy storage system.

[0041] To ensure the validity of the model, regular checks are conducted on the deviation between actual carbon emissions and the predicted model:

[0042] ΔC(t)=C actual (t)-C pred (t)

[0043] By analyzing the actual value of carbon emissions C actual (t) and the predicted value C pred (t), dynamically adjust the parameters in the model to improve prediction accuracy and control efficiency.

[0044] As a preferred solution of the method for constructing scenarios of factors affecting carbon emissions in the park described in the present invention, the modeling of each influencing factor also includes electric vehicle charging, energy storage and other electricity items in the park. The total power of electric vehicle charging is P EV , the carbon emission coefficient per kilowatt-hour is γ CO2 , then the contribution of electric vehicle charging to carbon emissions is:

[0045] E EV =P EV ×t×γ CO2

[0046] The energy storage system reduces carbon emissions during peak hours by shaving peak loads and filling valleys. The energy storage equipment stores electricity from renewable energy during valley hours and releases it during peak hours, thereby reducing the proportion of fossil energy use. Scenario modeling takes into account the charge and discharge management of energy storage equipment, real-time load demand and weather forecasts to ensure that charging and discharging are carried out at appropriate times to achieve the effect of optimizing carbon emissions. The carbon emission reduction benefit of the energy storage system is expressed as:

[0047] ΔE Storage =(E peak -E off-peak )×η storage

[0048] The energy consumption of other electrical equipment is also directly related to the carbon emissions of the park. Carbon emissions can be reduced by optimizing the operating time of other electrical equipment and reducing the power demand during peak hours. The total power of other electrical equipment is P other , the corresponding carbon emissions are:

[0049] E other =P other ×t×γ CO2

[0050] Where t represents the charging time.

[0051] As a preferred scheme of the method for constructing scenarios of factors affecting park carbon emissions described in the present invention, the optimized clustering results include selecting the optimal K value using the SC coefficient and the DBI parameter, and optimizing the clustering effect by minimizing 1 minus the sum of the SC coefficient and the DBI parameter.

[0052] A scenario construction system for factors affecting carbon emissions in a park, characterized by: including:

[0053] Build a scenario module, collect park data, and construct a scenario of carbon emission influencing factors based on the self-optimizing k-means++ clustering method.

[0054] The calculation module models and calculates each influencing factor.

[0055] Optimization module, calculates total carbon emissions and optimizes clustering results.

[0056] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0057] A computer-readable storage medium stores a computer program, which implements the steps of the method described above when executed by a processor.

[0058] Beneficial effects of the present invention: Through cluster analysis, the factors affecting carbon emissions in different functional spaces can be classified and analyzed in detail. The introduction of the Markov chain model can accurately simulate the indoor status and switching actions of people in different functional spaces, so as to more accurately predict and manage energy consumption changes. By storing energy during low periods and releasing energy during peak periods, energy consumption fluctuations in the park can be effectively smoothed. In addition, the intelligent control system can dynamically adjust the operation of equipment, combining personnel behavior prediction and real-time data to ensure that energy consumption and carbon emissions are minimized while meeting comfort levels. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them:

[0060] Figure 1 An overall flow chart of a method and system for constructing scenarios of factors affecting carbon emissions in a park provided in the first embodiment of the present invention.

[0061] Figure 2 A Markov chain model flow chart of a method and system for constructing scenarios of factors affecting carbon emissions in a park provided in the first embodiment of the present invention. DETAILED DESCRIPTION

[0062] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0063] Example 1, reference Figure 1 , Figure 2 , which is an embodiment of the present invention, provides a method for constructing a scenario of factors affecting carbon emissions in a park, including:

[0064] S1: Collect park data and construct scenarios of carbon emission influencing factors based on the self-optimizing k-means++ clustering method.

[0065] It should be noted that K-means is one of the top ten algorithms in the field of data mining. Its core idea is to cluster based on data similarity and measure the similarity of objects by distance. The smaller the distance, the higher the similarity, and the more likely it is to be assigned to the same cluster. However, traditional k-means also has some obvious defects: the random selection of the initial cluster center may lead to slower convergence and even affect the clustering effect. The number of clusters K needs to be set manually, and the optimal K value is often difficult to determine. In order to solve these problems, an improved self-optimizing k-means++ clustering method (SI-kmeans++) is used in the construction of the scenario of factors affecting carbon emissions in the park. The optimal K value is automatically found by combining the SC and DBI indicators, and the k-means++ method is used to optimize the initial cluster center. The specific implementation steps include finding the optimal K value, initializing the cluster center, classifying samples, and updating the cluster center.

[0066] Repeat sample classification and cluster center update until the cluster center no longer changes.

[0067] In the process of cluster center initialization, a sample is first randomly selected as the first cluster center, and then the shortest distance between each sample and the current existing center is calculated. Next, the next cluster center is selected according to the probability of each sample being selected as the next center (proportional to the square of the distance), until the number of selected cluster centers reaches K, and finally the cluster center initialization is ended. In the clustering process, the optimal K value is first determined, and then clustering is started according to the steps of cluster center initialization. Each sample is assigned to the cluster closest to it, and then the average value of the data in each cluster is calculated to update the cluster center. Then check whether the cluster center has changed. If it has changed, repeat the above steps. If there is no change, output the clustering result. In this process, the SC coefficient (silhouette coefficient) and DBI parameter (Davies-Bouldin index) are used to select the optimal K value, and the clustering effect is optimized by minimizing 1 minus the sum of the SC coefficient and the DBI parameter.

[0068] Furthermore, when this self-optimizing K-means++ algorithm is applied to the scenario construction of the factors affecting carbon emissions in the park, the algorithm is used to perform cluster analysis on the factors affecting carbon emissions in the park. First, when initializing the cluster center, a carbon emission influencing factor (such as the energy consumption of the HVAC system or lighting system) is randomly selected as the initial cluster center, and the distance between other influencing factors (such as occupancy rate, temperature, renewable energy input, etc.) and the center is calculated. Then, other cluster centers are selected based on the probability of the distance to ensure that the selected center can represent a variety of influencing factors.

[0069] In the clustering process, various influencing factors are grouped and clustered, which may include: indoor factors (such as occupancy, equipment switching action), time factors (such as year, month, season), weather factors (such as outdoor temperature, humidity, radiation) and calendar factors (such as holidays, weekends, etc.). By calculating the average carbon emission contribution value in each cluster, updating the cluster center, and checking whether the center has changed, the accuracy of the clustering results is ensured.

[0070] S2: Model and calculate each influencing factor.

[0071] It should be noted that in order to accurately model these behaviors, a Markov chain is used to describe the presence of people in the room. The Markov chain model regards the behavior of each individual as a movement process. For a single-person office, the presence of people in the room can be simulated through the sampling of a Markov chain. For a multi-person room, multiple Markov chains can be superimposed to simulate the overall occupancy of the room. In the simulation process, the event mechanism plays an important role. For example, commuting events and outing events can calculate the transition probability and average residence time of people between different states by assuming the time and duration of the event, thereby making more accurate predictions and management of the dynamic changes of indoor carbon emissions. Figure 2 Assume that the step length of staying in state i is ST i , then ST i It obeys the geometric distribution, expressed as:

[0072]

[0073] Among them, P{ST i =k} ​​represents the probability that an individual starts in state i and leaves state i after k steps and i steps.

[0074] The park is divided into multiple functional spaces, each of which has different sensitivities to different weather factors. According to the weights of each subsystem and influencing factors, a corresponding scenario model is constructed for each functional space:

[0075] It should be noted that in order to better simulate the impact of unpredictable factors on carbon emissions, we can divide the park into multiple functional spaces, each of which has different sensitivities to different weather factors. According to the weights of each subsystem and influencing factors, a corresponding scenario model is constructed for each functional space:

[0076] First, the relationship between weather factors and energy consumption is established based on the collected key weather data (temperature, humidity, rainfall, solar radiation, wind speed and direction) and building energy consumption data. Different functional spaces in the park (such as production buildings, office buildings, etc.) have different sensitivities to weather factors. Through weight analysis, sensitive factors can be identified and prediction models can be constructed. In order to achieve dynamic optimization, a real-time data feedback and prediction mechanism needs to be established, so that managers can adjust equipment operation according to weather changes and reduce energy waste. Machine learning models can predict weather changes and optimize the operation of air conditioning, heating and other equipment in advance. Intelligent control systems improve energy management efficiency through automated regulation. The results need to be continuously evaluated, and the model parameters are optimized by regularly checking the deviation of carbon emissions from model predictions. The coordinated scheduling of energy storage systems and functional spaces can smooth energy consumption and avoid sudden increases in carbon emissions. Through these methods, the park can effectively adjust equipment operation, reduce carbon emission fluctuations, and achieve sustainable development goals.

[0077] Establishing the correlation between weather factors and energy consumption:

[0078]

[0079] Among them, E i (t) represents the energy consumption of functional space i at time t, F j (t) represents the value of weather factor j at time t, w ij is the weight of function space i on weather factor j.

[0080] In order to achieve dynamic optimization of energy-consuming equipment, a real-time feedback and prediction mechanism is established, combined with a machine learning prediction model, so that the equipment can be adjusted in advance under different weather scenarios to avoid unnecessary energy consumption fluctuations, expressed as:

[0081] P(t+1)=α·P(t)+β·E pred (t+1)

[0082] Where P(t) represents the power demand of the device at time t, E pred (t+1) represents the energy consumption forecast for the future time point t+1 based on the weather forecast model, α represents the adjustment coefficient, and β represents the adjustment coefficient, which is used to balance the weight of historical data and predicted data.

[0083] For each functional space, calculate its carbon emissions:

[0084]

[0085] Among them, C i (t) represents the carbon emissions of functional space i at time t, η k represents the carbon emission coefficient of k energy subsystems, E ik (t) represents the energy consumption of the kth subsystem used by functional space i at time t.

[0086] In order to smooth the energy consumption of each functional space and avoid a sudden increase in carbon emissions, the coordinated scheduling of the energy storage system is introduced:

[0087] S(t)=S(t-1)+P charge (t)-P discharge (t)

[0088] Where S(t) represents the energy storage level of the energy storage system at time t, P charge (t) represents the charging power of the energy storage system, P discharge (t) represents the discharge power of the energy storage system.

[0089] To ensure the validity of the model, regular checks are conducted on the deviation between actual carbon emissions and the predicted model:

[0090] ΔC(t)=C actual(t)-C pred (t)

[0091] By analyzing the actual value of carbon emissions C actual (t) and the predicted value C pred (t), dynamically adjust the parameters in the model to improve prediction accuracy and control efficiency.

[0092] It should be noted that electric vehicle charging will significantly increase the total electricity demand of the park during peak load periods. The peak of electric vehicle charging usually occurs in the morning or evening on weekdays. By reasonably arranging the charging schedule and allocating charging to low-peak periods, the grid load and carbon emissions can be reduced.

[0093] Electric vehicle charging, energy storage and other electricity consumption items in the park. The total power of electric vehicle charging is P EV , the carbon emission coefficient per kilowatt-hour is γ CO2 , then the contribution of electric vehicle charging to carbon emissions is:

[0094]

[0095] The energy storage system reduces carbon emissions during peak hours by shaving peak loads and filling valleys. The energy storage equipment stores electricity from renewable energy during valley hours and releases it during peak hours, thereby reducing the proportion of fossil energy use. Scenario modeling takes into account the charge and discharge management of energy storage equipment, real-time load demand and weather forecasts to ensure that charging and discharging are carried out at appropriate times to achieve the effect of optimizing carbon emissions. The carbon emission reduction benefit of the energy storage system is expressed as:

[0096] ΔE Storage =(E peak -E off-peak )×η storage

[0097] The energy consumption of other electrical equipment is also directly related to the carbon emissions of the park. Carbon emissions can be reduced by optimizing the operating time of other electrical equipment and reducing the power demand during peak hours. The total power of other electrical equipment is P other , the corresponding carbon emissions are:

[0098] E other =P other ×t×γ CO2

[0099] Where t represents the charging time.

[0100] S3: Calculate total carbon emissions and optimize clustering results.

[0101] The SC coefficient and DBI parameter are used to select the optimal K value, and the clustering effect is optimized by minimizing 1 minus the sum of the SC coefficient and the DBI parameter.

[0102] The above embodiments also include a system for constructing scenarios of factors affecting carbon emissions in a park, specifically:

[0103] Build a scenario module, collect park data, and construct a scenario of carbon emission influencing factors based on the self-optimizing k-means++ clustering method.

[0104] The calculation module models and calculates each influencing factor.

[0105] Optimization module, calculates total carbon emissions and optimizes clustering results.

[0106] The computer device may be a server. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected via a system bus, and the communication interface is connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data cluster data of the power monitoring system. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for constructing a scenario of factors affecting carbon emissions in a park is implemented.

[0107] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0108] Example 2 is an embodiment of the present invention, which provides a method and system for constructing scenarios of factors affecting carbon emissions in a park. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through simulation experiments.

[0109] The goal of this experiment is to verify the effectiveness of clustering analysis of factors affecting carbon emissions in the park through the self-optimizing K-means++ clustering algorithm, and compare it with the traditional K-means clustering algorithm to demonstrate the advantages of the present invention in scenario construction and carbon emission optimization.

[0110] The experimental object is the factors affecting carbon emissions in an industrial park. The main parameters involved include temperature, humidity, occupancy rate, use of energy-consuming equipment, and energy consumption contribution of different functional spaces. The data source is the actual monitoring data of multiple buildings in the park, including production areas, office buildings, dormitories, and public areas. The experiment is divided into two stages: the experiment of the traditional K-means clustering algorithm and the experiment based on the self-optimizing K-means++ algorithm.

[0111] Existing technology (traditional K-means clustering algorithm) steps:

[0112] Select the main influencing factors of carbon emissions, including temperature, humidity, occupancy rate, equipment energy consumption, etc. as input data. Set the K value (i.e. the number of clusters) based on experience. Taking 3 clusters as an example, randomly select 3 initial cluster centers. By calculating the distance from the sample to the cluster center, assign each sample to the nearest cluster center. Calculate the average value of the data in each cluster and update the cluster center until the cluster center no longer changes. Finally, output the clustering results to determine the main influencing factors of carbon emissions under different categories.

[0113] Our invention steps:

[0114] The optimal K value is automatically found based on SC (silhouette coefficient) and DBI (Davies-Bouldin index), avoiding the deviation of manually setting the K value. The initial cluster center is optimized using the K-means++ method. This method selects representative initial center points by maximizing the distance between the sample and the cluster center, reducing the fluctuation of the clustering results. During the clustering process, the distance between each sample and the current cluster center is calculated, and the samples are assigned to the corresponding clusters according to the shortest distance. The average value of the carbon emission influencing factors of each cluster is calculated, and the cluster center is updated until the cluster center no longer changes. The final clustering results are output, and the accuracy of the clustering is adjusted through the optimization module to ensure that the carbon emission influencing factors of different functional spaces are clearly classified, which is convenient for subsequent carbon emission optimization.

[0115] The experimental results are shown in Table 1.

[0116] Table 1 Experimental results

[0117]

[0118] From the above experimental data, it can be seen that the self-optimizing K-means++ algorithm invented by us is significantly better than the traditional K-means algorithm in clustering carbon emissions. The self-optimizing K-means++ method automatically selects the optimal K value through the SC and DBI indicators, and combines the K-means++ method to optimize the initial clustering center, making the clustering center more representative, thereby improving the accuracy of the clustering results. Compared with the traditional method, the distribution of carbon emissions is more reasonable, avoiding the inaccurate clustering caused by local optimality.

[0119] From the data, it can be seen that the clustering results based on the self-optimization algorithm show that the carbon emissions are smoother. For example, the carbon emissions of clusters 1 and 3 are reduced by 50kg and 100kg respectively. This shows that by optimizing the clustering results, the utilization rate of energy-consuming equipment and the fluctuation of carbon emissions in the park are better controlled, which further verifies the practicality of the algorithm in the construction of carbon emission scenarios.

[0120] The self-optimizing K-means++ method better balances various influencing factors, such as the relationship between environmental factors such as temperature and humidity and personnel occupancy rate and equipment utilization rate. Through automated clustering optimization, it reduces the error of manual intervention and obtains a more accurate energy consumption prediction model.

[0121] The traditional K-means algorithm has a large randomness in the selection of the initial cluster center, which easily causes data fluctuations. For example, the carbon emissions of cluster 1 fluctuate greatly. The self-optimizing K-means++ algorithm significantly reduces this fluctuation, showing higher stability and accuracy, especially in dealing with complex and changeable park carbon emission factors.

[0122] In summary, our invention not only improves the accuracy of clustering through the self-optimizing K-means++ algorithm, but also realizes the dynamic optimization of carbon emission scenario construction, further improving the scientificity and effectiveness of the park's carbon management.

[0123] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for constructing scenarios of factors affecting carbon emissions in a park, characterized in that: include: Collect park data and construct scenarios of carbon emission influencing factors based on the self-optimizing k-means++ clustering method; Model and calculate each influencing factor; Calculate total carbon emissions and optimize clustering results.

2. The method for constructing scenarios of factors affecting park carbon emissions according to claim 1, characterized in that: The self-optimizing k-means++ clustering method includes automatically finding the best K value by combining the SC and DBI indicators, and using the k-means++ method to optimize the initial cluster center, and the steps include finding the best K value, initializing the cluster center, classifying samples, and updating the cluster center; Repeat sample classification and cluster center update until the cluster center no longer changes.

3. The method for constructing scenarios of factors affecting park carbon emissions as claimed in claim 2, characterized in that: The initial cluster center includes, in the process of cluster center initialization, randomly selecting a sample as the first cluster center, calculating the shortest distance between each sample and the current existing center; selecting the next cluster center according to the probability of each sample being selected as the next center, until the number of selected cluster centers reaches K, and finally ending the cluster center initialization; In the clustering process, the optimal K value is first determined, and clustering is started according to the steps of cluster center initialization; each sample is assigned to the cluster closest to it, the average value of the data in each cluster is calculated, and the cluster center is updated; Check whether the cluster center has changed. If so, repeat the cluster center initialization and clustering process. Otherwise, output the clustering result.

4. The method for constructing scenarios of factors affecting park carbon emissions as claimed in claim 3, characterized in that: The construction of the carbon emission influencing factor scenario based on the self-optimizing k-means++ clustering method includes applying the self-optimizing K-means++ algorithm to the scenario construction of the carbon emission influencing factors of the park, and performing cluster analysis on the carbon emission influencing factors of the park through the algorithm; When initializing the cluster center, randomly select a carbon emission influencing factor as the initial cluster center, calculate the distance between other influencing factors and the center, and select other cluster centers based on the probability of the distance to ensure that the selected center can represent a variety of influencing factors; In the clustering process, various influencing factors are grouped and clustered, including: indoor factors, time factors, weather factors, and calendar factors; The accuracy of clustering results is ensured by calculating the average carbon emission contribution value in each cluster, updating the cluster center, and checking whether the center has changed.

5. The method for constructing scenarios of factors affecting park carbon emissions as claimed in claim 4, characterized in that: The modeling and calculation of each influencing factor includes: indoor factors use Markov chain to describe the indoor situation of people. For a single-person office, the indoor situation of people is simulated by sampling a Markov chain; for a multi-person room, multiple Markov chains are superimposed to simulate the overall occupancy of the room; assuming that the step length staying in state i is ST i , then ST i It obeys the geometric distribution, expressed as: Among them, P{ST i =k} ​​represents the probability that an individual starts in state i and leaves state i after k steps and i steps; The park is divided into multiple functional spaces, each of which has different sensitivities to different weather factors. According to the weights of each subsystem and influencing factors, a corresponding scenario model is constructed for each functional space: Based on the collected key weather data and building energy consumption data, the relationship between weather factors and energy consumption is established; different functional spaces in the park have different sensitivities to weather factors, and sensitive factors are identified through weight analysis and a prediction model is constructed; Establishing the correlation between weather factors and energy consumption: Among them, E i (t) represents the energy consumption of functional space i at time t, F j (t) represents the value of weather factor j at time t, w ij is the weight of functional space i on weather factor j; In order to achieve dynamic optimization of energy-consuming equipment, a real-time feedback and prediction mechanism is established, combined with a machine learning prediction model, so that the equipment can be adjusted in advance under different weather scenarios to avoid unnecessary energy consumption fluctuations, expressed as: P(t+1)=α·P(t)+β·E pred (t+1) Where P(t) represents the power demand of the device at time t, E pred (t+1) represents the energy consumption forecast for the future time point t+1 based on the weather forecast model, α represents the adjustment coefficient, and β represents the adjustment coefficient, which is used to balance the weight of historical data and forecast data; For each functional space, calculate its carbon emissions: Among them, C i (t) represents the carbon emissions of functional space i at time t, η k represents the carbon emission coefficient of k energy subsystems, E ik (t) represents the energy consumption of the kth subsystem used by functional space i at time t; In order to smooth the energy consumption of each functional space and avoid a sudden increase in carbon emissions, the coordinated scheduling of the energy storage system is introduced: S(t)=S(t-1)+P charge (t)-P discharge (t) Where S(t) represents the energy storage level of the energy storage system at time t, P charge (t) represents the charging power of the energy storage system, P discharge (t) represents the discharge power of the energy storage system; To ensure the validity of the model, regular checks are conducted on the deviation between actual carbon emissions and the predicted model: ΔC(t)=C actual (t)-C pred (t) By analyzing the actual value of carbon emissions C actual (t) and the predicted value C pred (t), dynamically adjust the parameters in the model to improve prediction accuracy and control efficiency.

6. The method for constructing scenarios of factors affecting park carbon emissions as claimed in claim 5, characterized in that: The modeling of each influencing factor also includes electric vehicle charging, energy storage and other electricity items in the park. The total power of electric vehicle charging is P EV , the carbon emission coefficient per kilowatt-hour is γ CO2 , then the contribution of electric vehicle charging to carbon emissions is: The energy storage system reduces carbon emissions during peak hours by shaving peak loads and filling valleys. The energy storage equipment stores electricity from renewable energy during valley hours and releases it during peak hours, thereby reducing the proportion of fossil energy use. Scenario modeling takes into account the charge and discharge management of energy storage equipment, real-time load demand and weather forecasts to ensure charging and discharging at appropriate times to achieve the effect of optimizing carbon emissions. The carbon emission reduction benefit of the energy storage system is expressed as: ΔE Storage =(E peak -E off-peak )×η storage The energy consumption of other electrical equipment is also directly related to the carbon emissions of the park. Carbon emissions can be reduced by optimizing the operating time of other electrical equipment and reducing the power demand during peak hours. The total power of other electrical equipment is P other , the corresponding carbon emissions are: AND other =P other ×t×γ CO2 Where t represents the charging time.

7. The method for constructing scenarios of factors affecting park carbon emissions according to claim 6, characterized in that: The optimization of clustering results includes selecting an optimal K value using the SC coefficient and the DBI parameter, and optimizing the clustering effect by minimizing 1 minus the sum of the SC coefficient and the DBI parameter.

8. A system for constructing scenarios of factors affecting park carbon emissions using the method described in any one of claims 1 to 7, characterized in that: Build a scenario module, collect park data, and build a scenario of carbon emission influencing factors based on the self-optimizing k-means++ clustering method; The calculation module models and calculates each influencing factor; Optimization module, calculates total carbon emissions and optimizes clustering results.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images