Method for predicting and correcting carbon emission in construction process

Abstract

The invention provides a method for predicting and correcting carbon emission in a construction process, which belongs to the technical field of carbon emission prediction and comprises the following steps of: monitoring a total error value between a carbon emission predicted value and a carbon emission actual value in real time in the construction process; and when the total error value is greater than a first set value, calibration is carried out, and weighted modeling, nonlinear correction and local optimization are carried out on the total error value in sequence. And optimizing model parameters of the carbon emission prediction model and correction parameters introduced by nonlinear correction by using a genetic algorithm, and substituting the optimized model parameters and correction parameters into the carbon emission prediction model. The prediction error of the optimized carbon emission prediction model is compared with the first set value again, and if the prediction error is still larger than the first set value, the calibration process is repeated till the prediction error is smaller than the first set value. On the basis of the scheme, the first carbon emission prediction model can be adjusted in real time according to the real situation encountered in the actual construction process, and the accuracy of the prediction result in the face of accidental situations is improved.

Description

Technical Field

[0001] The present invention generally relates to the technical field of carbon emission prediction, and particularly to a method for predicting and correcting carbon emissions during the construction process. Background Art

[0002] In the prior art, the prediction of the total carbon emissions during the construction process is carried out by using a multiple regression model. Historical data is used in advance to train the multiple regression model. The training process includes: using historical data as input, historical carbon emission values as label data, and predicting the total carbon emissions as output for multiple rounds of training. The historical data includes external environmental factors, construction site characteristics, and dynamic change factors. Then, during the actual construction process, the actual values of the above data are input into the multiple regression model in real time to predict the total carbon emissions.

[0003] However, this prediction method is based on the historical data of past construction processes. Accidental situations in the actual construction process and special situations not encountered in the historical data are not considered in the prediction process of the total carbon emissions. This leads to an error between the total carbon emissions predicted by inputting various data in the construction process into the multiple regression model and the actual total carbon emissions. When the error is too large, the carbon emission prediction result loses its guiding significance for subsequent adjustment of the construction plan.

[0004] Therefore, it is necessary to monitor and correct the accuracy of the predicted value of the carbon emissions during the current construction process according to the actual situations encountered during the current construction process. Summary of the Invention

[0005] In view of the above-mentioned defects or deficiencies in the prior art, it is desirable to provide a method for predicting and correcting carbon emissions during the construction process.

[0006] The present invention provides a method for predicting and correcting carbon emissions during the construction process, including: Obtaining a first carbon emission prediction value and an actual carbon emission value during the period from the first moment to the second moment during the construction process, and calculating a first total error between the first carbon emission prediction value and the actual carbon emission value; the first carbon emission prediction value is predicted by a first carbon emission prediction model; If the first total error is greater than a first set value, constructing a weighted relationship between the first total error and multiple influencing factors; Substituting the first total error into the weighted relationship, and sequentially performing non-linear correction and local optimization on the weighted relationship to obtain an optimized weighted relationship; a correction parameter is introduced during the non-linear correction process; According to the optimized weighted relationship, using a genetic algorithm to optimize the first model parameters of the first carbon emission prediction model and the correction parameter to obtain optimized model parameters and optimized correction parameters; An error term is introduced into the first carbon emission prediction model, the optimized model parameters are substituted into the first carbon emission prediction model, and the optimized weighted relationship formula with the optimized correction parameters is substituted into the error term to obtain a second carbon emission prediction model.

[0007] According to the technical solution provided by the present invention, after obtaining the second carbon emission prediction model, it further includes: Obtain the second carbon emission prediction value during the period from the first moment to the second moment during the construction process, and calculate the total second error between the second carbon emission prediction value and the actual carbon emission value; the second carbon emission prediction value is predicted by the second carbon emission prediction model; If the total second error is greater than the first set value, the second carbon emission prediction model is cyclically optimized according to the total second error until the total second error between the second carbon emission prediction value predicted by using the second carbon emission prediction model and the actual carbon emission value is less than or equal to the first set value.

[0008] According to the technical solution provided by the present invention, the first carbon emission prediction model is obtained in the following manner: Obtain historical data; the historical data includes multiple influencing factors during past construction processes; Obtain the carbon emission values during past construction processes to obtain historical carbon emission values; Obtain a multiple regression model; Use multiple influencing factors during past construction processes as the input of the multiple regression model, the historical carbon emission values as label data, and the first carbon emission prediction value as the output, and train the multiple regression model to obtain the first carbon emission prediction model.

[0009] According to the technical solution provided by the present invention, the first set value satisfies Formula 1: Formula 1; wherein, represents the first set value, T 总 represents the total construction duration, C 0 represents the carbon emission standard value for the entire construction process, t’ represents the monitoring duration from the first moment to the second moment, z represents the error tolerance threshold.

[0010] According to the technical solution provided by the present invention, the weighted relationship formula is; Formula 2; wherein, represents the weighted relationship formula, and its value is equal to the total first error, Indicates the weight of the error caused by the i th influencing factor, Indicates the error caused by the i th influencing factor, Indicates the non - linear influence of the error caused by the i th influencing factor, n Indicates the total number of influencing factors, i = 1, 2, 3... n .

[0011] According to the technical solution provided by the present invention, using Formula Three, perform non - linear correction on the weighted relational expression to obtain an expression for the total corrected error; Formula Three; Wherein, Indicates the carbon emission prediction value after non - linear correction, Indicates the first carbon emission prediction value, Indicates the correction coefficient, Indicates the non - linear correction factor, Indicates the weighted relational expression, Indicates the expression for the total corrected error; The correction parameters include the correction coefficient and the non - linear correction factor.

[0012] According to the technical solution provided by the present invention, perform local optimization on the expression for the total corrected error to obtain the optimized weighted relational expression; Local optimization includes: Obtain the actual carbon emission sub - values generated by each influencing factor during the period from the first moment to the second moment; Obtain the error sub - values caused by each influencing factor; Calculate the ratios of the error sub - values of multiple said influencing factors to the actual carbon emission sub - values to obtain multiple first ratios; According to multiple said first ratios, re - allocate weights for multiple influencing factors to obtain optimized weights; Calculate the mean value of multiple said first ratios to obtain the first mean value; Take the influencing factors corresponding to multiple first ratios greater than the first mean value as the local optimization targets; Use the genetic algorithm to optimize the correction coefficient and the non - linear correction factor of the local optimization targets to obtain an optimized correction coefficient and an optimized non - linear correction factor; Substitute the optimized correction coefficient, the optimized non - linear correction factor, and the optimized weights into Formula Two to obtain the optimized weighted relational expression.

[0013] According to the technical solution provided by the present invention, based on the optimized weighting relationship, the genetic algorithm is used to optimize the first model parameters of the first carbon emission prediction model and the correction parameters to obtain optimized model parameters and optimized correction parameters, including: Set algorithm parameters, fitness function, and termination conditions; the algorithm parameters include crossover probability and mutation probability; Before the genetic algorithm optimization iteration, update the values of the crossover probability and the mutation probability according to the first total error value; During the genetic algorithm optimization iteration, make the crossover probability and the mutation probability decrease as the number of iterations increases; After the genetic algorithm reaches the termination condition, optimize to obtain the historical optimal model parameters and the historical optimal correction parameters; Under the condition that the loss function satisfies gradient descent, adjust the historical optimal model parameters to obtain the optimized model parameters; Use the historical optimal correction parameters as the optimized correction parameters.

[0014] According to the technical solution provided by the present invention, updating the values of the crossover probability and the mutation probability according to the first total error value includes: Obtain the initial values of the crossover probability and the mutation probability to get the initial crossover probability and the initial mutation probability; Calculate the ratio of the first total error value to the actual carbon emission value to get the second ratio; Make the value of the crossover probability equal to the initial crossover probability plus the product of the initial crossover probability and the second ratio; Make the value of the mutation probability equal to the initial mutation probability plus the product of the initial mutation probability and the second ratio.

[0015] According to the technical solution provided by the present invention, the genetic algorithm also has an evolutionary equation with a non - linear penalty function, as shown in Formula Four: Formula Four; Wherein, represents the individual after the genetic algorithm in the t +1 generation, represents the individual of the current t generation, t represents the number of current iterations, and V are both adjustment factors for adjusting the evolutionary step size, represents the adaptive fitness function, represents the gradient of the adaptive fitness function, represents the random perturbation factor, represents the random perturbation term, G represents the individual of the genetic algorithm.

[0016] The beneficial effects of the present invention are as follows: During the construction process, the total first error between the first carbon emission prediction value of the first carbon emission prediction model and the actual carbon emission value is monitored in real time. When the total first error is greater than the first set value, calibration is performed, including: performing weighted modeling on the total first error in sequence (i.e., constructing a weighted relationship between the total first error and multiple influencing factors), non-linear correction, and local optimization. The genetic algorithm is used to optimize the first model parameters of the first carbon emission prediction model and the correction parameters introduced by the non-linear correction, and the optimized model parameters and correction parameters are substituted into the first carbon emission prediction model. The prediction error of the optimized first carbon emission prediction model is compared with the first set value again. If it is still greater than the first set value, the calibration process is repeated until the prediction error is less than the first set value. Based on the above solution, the first carbon emission prediction model can be adjusted in real time according to the actual situation encountered during the construction process, improving the accuracy of the carbon emission prediction result in the face of accidental situations. Description of the Drawings

[0017] Other features, objects, and advantages of the present invention will become more apparent by reading the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 It is a schematic flowchart of a method for predicting and correcting carbon emissions during the construction process. Detailed Embodiments

[0018] The present invention will be further described in detail below with reference to the drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the related invention and are not intended to limit the invention. Additionally, it should be noted that only the parts related to the invention are shown in the drawings for the convenience of description.

[0019] It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the drawings and embodiments.

[0020] Please refer to Figure 1 , the present invention provides a method for predicting and correcting carbon emissions during the construction process, including: S1: Obtain the first carbon emission prediction model, including: S1-1: Obtain historical data; the historical data includes multiple influencing factors during previous construction processes; S1-2: Obtain the carbon emission values during previous construction processes to obtain historical carbon emission values; S1-3: Obtain a multiple regression model; S1-4: Use multiple influencing factors in previous construction processes as the input of the multiple regression model, the historical carbon emission value as the label data, and the first carbon emission prediction value as the output to train the multiple regression model to obtain the first carbon emission prediction model.

[0021] Specifically, the first carbon emission prediction model is used to input multiple influencing factors at the first moment and predict the carbon emission value during the construction process from the first moment to the second moment; the first carbon emission prediction model has first model parameters; the multiple influencing factors include external environment factors, construction site characteristic factors, and dynamic change factors.

[0022] Among them, the external environment factors: the temperature and humidity of the construction site; The construction site characteristic factors: the type of construction equipment and the amount of material used; The dynamic change factors: construction efficiency and working hours.

[0023] The multiple regression model (the first carbon emission prediction model) is shown in Formula VIII: Formula VIII.

[0024] Therefore, the relationship between the first carbon emission prediction value and the actual carbon emission value can be expressed by Formula IX: Formula IX; Among them: represents the actual carbon emission value, represents the first carbon emission prediction value, X is the external environment factor; such as temperature, humidity, etc., W is the construction site characteristic factor; such as equipment type, material use, etc. When calculating specifically, the specific value of the equipment type is the carbon emission value generated by the normal operation of the equipment per unit time; the material use represents the carbon emission value generated by consuming a unit mass of this material; Z is the dynamic change factor during the construction process; such as the efficiency of the operators, working hours, etc. The efficiency is the ratio of the actual workload completed by the operators per unit time to the expected workload; β 0 is the constant term, β 1, β 2, β 3 are the regression coefficients of each factor, that is, the model parameters; for example, in the first carbon emission prediction model, they are the first model parameters; is the weighted relationship formula.

[0025] S2: Obtain the first carbon emission prediction value and the actual carbon emission value during the period from the first moment to the second moment in the construction process; the first carbon emission prediction value is predicted by the first carbon emission prediction model; Calculate the total first error value between the first carbon emission prediction value predicted by using the first carbon emission prediction model during the period from the first moment to the second moment in the construction process and the actual carbon emission value generated from the first moment to the second moment during construction, including: S2-1: Obtain multiple influencing factors at the first moment; S2-2: Input the multiple influencing factors at the first moment into the first carbon emission prediction model to predict the first carbon emission prediction value; S2-3: Obtain the actual carbon emission value generated from the first moment to the second moment during construction; Specifically, various types of sensors at the actual construction site measure different data and calculate the actual values of each factor; this calculation process belongs to the prior art.

[0026] S2-4: Calculate the absolute value of the difference between the first carbon emission prediction value and the actual carbon emission value; S2-5: Take the absolute value of the difference as the total first error value.

[0027] Specifically, the first moment and the second moment of the present invention are set to be spaced 5-10 minutes apart; thus, when an accidental situation occurs, the accidental situation can be taken into account in the influence on carbon emission prediction within a relatively short time; moreover, it will not increase the data operation amount due to overly frequent detection and correction.

[0028] S3: Judge whether the total first error value is greater than the first set value; If the total first error value is greater than the first set value, perform steps S5 to S10; otherwise, perform step S4; Furthermore, the first set value satisfies formula one: Formula one; Wherein, represents the first set value, T 总 represents the total construction duration, C 0 represents the carbon emission standard value of the entire construction process, t’ represents the monitoring duration from the first moment to the second moment, z represents the error tolerance threshold.

[0029] Among them, the carbon emission standard value is artificially set according to the specified emission standard. The size of the error tolerance threshold is also artificially set. The larger it is, the larger the error between the predicted value and the actual value of the first carbon emission prediction model is allowed; on the contrary, a smaller error is allowed, which can make the prediction of the calibrated first carbon emission prediction model more accurate. In the present invention, the error tolerance threshold is set to 1; thus, the prediction error of the calibrated first carbon emission prediction model will be limited within 1%.

[0030] S4: Update the first moment and the second moment, and repeat steps S2 to S3; Specifically, the way to update the first moment and the second moment is as follows: After the monitoring duration has passed, the current moment is taken as the first moment, and the moment after another monitoring duration is taken as the second moment.

[0031] Suppose the monitoring duration is set to 1 minute, and the first monitoring time is 8:20 am. Then the first moment is the moment at 8:20 am sharp, and the second moment is the moment at 8:21 am sharp.

[0032] After updating the first moment and the second moment, the second monitoring time is 8:21 am. Then the first moment is the moment at 8:21 am sharp, and the second moment is the moment at 8:22 am sharp.

[0033] S5: Construct a weighted relationship (i.e., perform weighted modeling) between the total first error value and multiple influencing factors; Substitute the total first error value into the weighted relationship, and perform non-linear correction and local optimization on the weighted relationship in turn to obtain an optimized weighted relationship; a correction parameter is introduced in the non-linear correction process; At the same time, an error term also needs to be introduced for the first carbon emission prediction model; Furthermore, in order to more accurately capture the error sources, a method of weighted modeling of errors is introduced. The total error value is divided into the sum of the errors brought by each of the multiple influencing factors, so that the error can more truly reflect the current construction situation.

[0034] Using Formula 2, perform weighted modeling on the total first error value to obtain a weighted relationship; Formula 2; Among them, represents the weighted relationship, and its value is equal to the total first error value, represents the weight of the error brought by the i th influencing factor, represents the i th influencing factor brings the error, represents thei The non - linear influence of the error brought by each influencing factor n Indicates the total number of influencing factors i = 1, 2, 3... n 。

[0035] Furthermore, in order to cope with the unforeseen non - linear errors during the construction process, a non - linear correction and local optimization mechanism is also introduced.

[0036] Using Equation Three, perform non - linear correction on the weighted relationship formula to obtain the expression of the total corrected error value; Equation Three; Wherein, Indicates the carbon emission prediction value after non - linear correction Indicates the first carbon emission prediction value Indicates the correction coefficient Indicates the non - linear correction factor Indicates the weighted relationship formula Indicates the expression of the total corrected error value; The non - linear correction factor is usually less than 1, which can control the intensity of error correction, making the correction more adaptable to the actual situation and avoiding over - adjustment of errors.

[0037] The correction parameters include the correction coefficient and the non - linear correction factor.

[0038] Specifically, by introducing dynamic weighting and non - linear influence factors, the model can adjust according to the contribution degree of each error to the prediction error, thereby improving the fitting effect of the model. This error adjustment method enables the regression model to more flexibly adapt to the complex and changeable environmental conditions during the construction process, improving the accuracy and reliability of prediction, because it can accurately reflect the specific influence of different errors on the final prediction result.

[0039] Furthermore, perform local optimization on the expression of the total corrected error value to obtain the optimized weighted relationship formula; Local optimization includes: Obtain the actual carbon emission sub - values generated by each influencing factor during the period from the first moment to the second moment; The specific obtaining method is to detect the specific values of each influencing factor and substitute them into the corresponding formula to calculate the carbon emission value corresponding to the influencing factor. The formula calculation method belongs to the prior art; Obtain the error sub - values brought by each influencing factor; Calculate the ratio of the error sub - values of multiple said influencing factors to the actual carbon emission sub - values to obtain multiple first ratios; According to multiple said first ratios, re - allocate weights to multiple influencing factors to obtain optimized weights; Specifically, the way of re - allocating weights is: The weights are allocated according to the magnitudes of their respective corresponding first ratios. The larger the first ratio, the greater the reallocated weight. At the same time, ensure that the sum of all weights is equal to 1.

[0040] For example, first calculate the sum of all the first ratios; then let the optimized weights corresponding to the multiple influencing factors be equal to the ratios of the first ratios corresponding to each influencing factor to the sum of all the first ratios.

[0041] The local optimization further includes: Calculate the mean value of the multiple first ratios to obtain the first mean value; Take the influencing factors corresponding to the multiple first ratios greater than the first mean value as the local optimization targets; the purpose here is to screen out the influencing factors that have a greater impact on the error.

[0042] Use the genetic algorithm to optimize the correction coefficient and the non - linear correction factor of the local optimization target to obtain the optimized correction coefficient and the optimized non - linear correction factor; Substitute the optimized correction coefficient, the optimized non - linear correction factor, and the optimized weight into Formula Two to obtain the optimized weighted relationship formula.

[0043] Specifically, for the genetic algorithm applied in the local optimization process, the variable range is set to the range of the correction coefficient and the non - linear correction factor; The loss function is set as: Formula Five; Wherein, θ represents the variable, which is the correction coefficient and the non - linear correction factor in this embodiment, represents the loss function, represents the influencing factor i at the weight of the sensor j ; represents the j th factor detected by the th sensor, and the actual carbon emission value obtained after calculation, that is, the real carbon emission data obtained on - site, θ represents the predicted carbon emission value of the j th observation sample calculated based on the variable i = 1, 2, 3... n , n represents the total number of influencing factors, j = 1, 2, 3... m , m represents the total number of sensors.

[0044] The remaining implementation manners are the same as those in step S6.

[0045] Specifically, the main optimization objective of the genetic algorithm is to correct the coefficients γ and the non-linear correction factor α . These parameters directly affect the correction strength and correction method of the error. By simulating the process of natural selection, the genetic algorithm gradually iteratively optimizes these parameters to find the parameter combination that is most suitable for the current error pattern, so as to improve the accuracy and stability of the prediction model.

[0046] The criterion for the optimal solution is to minimize the error between the corrected carbon emission prediction value and the actual carbon emission value. Specifically, the optimization process measures the quality of the solution by minimizing the loss function (usually the sum of the squares of the prediction errors). Ultimately, the optimal solution will be those parameter combinations that minimize the error and have the strongest model stability under certain constraints.

[0047] During the construction process, certain factors or errors (such as equipment errors, external environmental factors, etc.) have a greater impact on the predicted carbon emission values. Therefore, in these local areas, the errors will be particularly prominent. By analyzing the current error distribution, the system identifies the local areas or factors with large error changes and focuses on optimizing the parameters in these areas. By dynamically monitoring the changes at the construction site, the algorithm can identify the key parts that need local optimization, thus achieving more accurate error correction.

[0048] S6: According to the optimized weighting relationship, use the genetic algorithm to optimize the first model parameters of the first carbon emission prediction model and the correction parameters to obtain optimized model parameters and optimized correction parameters; Among them, the first model parameters are those mentioned above β 0, β 1, β 2, β 3; the correction parameters include those mentioned above and ; Specifically, it includes: Set the algorithm parameters, fitness function, and termination conditions; the algorithm parameters include the crossover probability and mutation probability; Among them, the algorithm parameters include: Population size, chromosome length, crossover probability, mutation probability, selection operator, variable range (in this step, it is the range of model parameters and correction parameters), and maximum number of iterations; In this embodiment, the population size is set to: 100; The chromosome length is: 30 - 50; The initial value of the crossover probability is: 0.8; The initial value of the mutation probability is: 0.05; The maximum number of iterations is any value between 100 and 1000; in this embodiment, it is set to 200.

[0049] Fitness function:

[0050] Formula Six; Wherein, represents the fitness function, and both represent the weight factors in the fitness function, represents the error change weight factor, represents the change in error compared to the previous iteration, represents the monitoring duration, represents the actual carbon emission value, represents the first carbon emission prediction value, m is the number of sensors.

[0051] Ending condition: The minimum value of the fitness function no longer decreases after consecutive iterations and the maximum number of iterations is reached.

[0052] Furthermore, before the genetic algorithm optimization iteration, according to the total first error, updating the values of the crossover probability and the mutation probability includes: Obtaining the initial values of the crossover probability and the mutation probability to get the initial crossover probability and the initial mutation probability; Calculating the ratio of the total first error to the actual carbon emission value to get a second ratio; Making the value of the crossover probability equal to the initial crossover probability plus the product of the initial crossover probability and the second ratio; Making the value of the mutation probability equal to the initial mutation probability plus the product of the initial mutation probability and the second ratio.

[0053] Specifically, generally the proportion of the general error is small, and the increased crossover probability and mutation probability will not be greater than 1; in some special cases, when it is greater than 1 after increase, the crossover probability and the mutation probability are set to 1.

[0054] In each iteration of the genetic algorithm, with the change of the external environment, the system will calculate the change of the prediction error. If the error increases (for example, abnormal error caused by equipment failure or material change), it indicates that the current search strategy of the system may not fully explore the problem space, and the crossover probability and the mutation probability can be appropriately increased. The crossover operation helps to explore the new solution space, while the mutation operation increases the diversity of the solutions, thus preventing the algorithm from falling into a local optimal solution.

[0055] Furthermore, during the genetic algorithm optimization iteration, making the crossover probability and the mutation probability decrease as the number of iterations increases; The specific reduction method is as follows: a fixed value is reduced in each generation. The fixed value of the crossover probability is equal to the ratio of the initial crossover probability to the maximum number of iterations. The fixed value of the mutation probability is equal to the ratio of the initial mutation probability to the maximum number of iterations.

[0056] If the external environment changes drastically (for example, the equipment at the construction site suddenly fails or the materials change), it may cause a significant increase in the carbon emission prediction error. Based on the above settings, in this case, the system will automatically enhance the intensity of the mutation operation. By increasing the mutation probability, the algorithm can jump out of the local area of the current solution in a short time and search for a more suitable solution to adapt to the new environment. The intensity of the mutation operation can be adjusted according to the change amplitude of the error: when the error changes greatly, increase the mutation rate to enhance the search ability; when the error changes smoothly, reduce the mutation rate to reduce the risk of over-search.

[0057] Furthermore, after the genetic algorithm reaches the end condition, the historical optimal model parameters and historical optimal correction parameters are obtained through optimization; Under the condition that the loss function satisfies gradient descent, the historical optimal model parameters are adjusted to obtain the optimized model parameters; The gradient descent formula is as shown in Formula Seven: Formula Seven; Among them, θ t represents the current variable, θ t+1 represents the updated variable, represents the learning rate, represents the gradient of the loss function with respect to the variable.

[0058] Finally, the historical optimal correction parameters are used as the optimized correction parameters.

[0059] Specifically, through multiple iterations, all weights will be gradually adjusted according to the current error. Each time an update is made, all model parameters of the regression model will be adjusted according to the calculated gradient to minimize the loss function and gradually make the prediction results closer to the actual values. This process is progressive. After multiple updates, the model weights will converge to the optimal solution, thereby improving the prediction accuracy.

[0060] Furthermore, the genetic algorithm also has an evolutionary equation with a non-linear penalty function, as shown in Formula Four: Formula Four; Among them, represents the individual after the genetic algorithm in t the +1 generation, that is, the new generation of individuals or optimized parameters after this round of evolution and optimization, Represents an individual in the current t-th generation, that is, the solution being optimized by the current genetic algorithm. t Represents the number of current iterations. and V Both are adjustment factors for adjusting the evolutionary step size. Represents the adaptive fitness function. Represents the gradient of the adaptive fitness function. Represents the random perturbation factor, which is used to avoid premature convergence. Represents the random perturbation term. G Represents an individual of the genetic algorithm, that is, the solution vector involved in the optimization process of the genetic algorithm.

[0061] Specifically, applying the evolutionary equation of the nonlinear penalty function can further increase the complexity of the algorithm, enabling it to be closer to the actual construction situation and improving the optimization ability at the same time.

[0062] S7: Substitute the optimized model parameters into the first carbon emission prediction model, and substitute the optimized weighted relationship with the optimized correction parameters into the error term to obtain the second carbon emission prediction model, referring to Formula Ten: Formula Ten; Among them, Represents the second carbon emission prediction model. β 0, β 1, β 2, β 3 all represent optimized model parameters; Represents the optimized weight; The optimized correction parameters include: the optimized correction coefficient and the optimized nonlinear correction factor .

[0063] After the above steps, the second carbon emission prediction model can be obtained. Since the model parameters of this model are optimized by the algorithm, there is still a low probability that the second carbon emission model still cannot accurately predict the carbon emission value. To exclude this small probability event, after obtaining the second carbon emission prediction model, it also includes: S8: Obtain the second carbon emission prediction value during the period from the first moment to the second moment in the construction process; the second carbon emission prediction value is predicted by the second carbon emission prediction model; Calculate the total second error between the second carbon emission prediction value predicted by using the second carbon emission prediction model and the actual carbon emission value during the period from the first moment to the second moment in the construction process; S9: Judge: Whether the total second error is greater than the first set value; If the total second error is greater than the first set value, the second carbon emission prediction model is cyclically optimized according to the total second error until the total second error between the second carbon emission prediction value obtained by predicting using the second carbon emission prediction model and the actual carbon emission value is less than or equal to the first set value; then proceed to step S10; The cyclic optimization specifically includes: taking the total second error as the total first error; taking the current model parameters as the first model parameters, and then repeating steps S5 and S9; If the total second error is less than or equal to the first set value, directly proceed to step S10.

[0064] S10: Use the second carbon emission prediction model finally obtained through cyclic optimization to predict the carbon emission value during the construction process; and return to step S4.

[0065] Specifically, after using the second carbon emission prediction model to complete the prediction of the carbon emission value, returning to step S4 can repeatedly monitor during the entire construction process whether the currently used carbon emission prediction model can still accurately predict the carbon emission value. Once the actual situation changes, resulting in too large an error between the actual carbon emission value and the predicted carbon emission value, the second carbon emission prediction model that has been cyclically optimized still needs to be optimized again to enable it to adapt to the new actual situation and ensure the prediction accuracy.

[0066] Specifically, after steps S7 to S10, in order to avoid the situation that after the optimized correction parameters and optimized model parameters obtained by genetic algorithm optimization are substituted into the first carbon emission prediction model, the error still cannot be less than the first set value, the present invention also sets a cyclic optimization mechanism until after the optimized correction parameters and optimized model parameters are substituted into the first carbon emission prediction model, the predicted error is less than the first set value.

[0067] Based on this mechanism, it can be ensured that the finally optimized result meets the requirements.

[0068] The above description is only the preferred embodiments of the present invention and the description of the applied technical principles. Those skilled in the art should understand that the scope of the invention involved in the present invention is not limited to the technical solutions formed by the specific combination of the above technical features, and should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the inventive concept. For example, the technical solutions formed by mutually substituting the above features with the technical features (but not limited to) having similar functions disclosed in the present invention.

Claims

1. A carbon emission prediction and correction method for the construction process, characterized in that Including: Obtain the first carbon emission prediction value and the actual carbon emission value during the period from the first moment to the second moment in the construction process, and calculate the total first error between the first carbon emission prediction value and the actual carbon emission value; The first carbon emission prediction value is predicted by a first carbon emission prediction model; If the total first error is greater than a first set value, construct a weighted relationship between the total first error and multiple influencing factors; Substitute the total first error into the weighted relationship, and perform non-linear correction and local optimization on the weighted relationship in sequence to obtain an optimized weighted relationship; the non-linear correction process introduces a correction parameter; According to the optimized weighted relationship, use a genetic algorithm to optimize the first model parameters of the first carbon emission prediction model and the correction parameter to obtain optimized model parameters and optimized correction parameters; Introduce an error term into the first carbon emission prediction model, substitute the optimized model parameters into the first carbon emission prediction model, and substitute the optimized weighted relationship with the optimized correction parameter into the error term to obtain a second carbon emission prediction model.

2. The carbon emission prediction and correction method during construction according to claim 1, wherein After obtaining the second carbon emission prediction model, it further includes: Obtain the second carbon emission prediction value during the period from the first moment to the second moment in the construction process, and calculate the total second error between the second carbon emission prediction value and the actual carbon emission value; the second carbon emission prediction value is predicted by the second carbon emission prediction model; If the total second error is greater than the first set value, perform cyclic optimization on the second carbon emission prediction model according to the total second error until the total second error between the second carbon emission prediction value predicted by using the second carbon emission prediction model and the actual carbon emission value is less than or equal to the first set value.

3. A carbon emission prediction and correction method for construction processes according to claim 1, characterized in that, The first carbon emission prediction model is obtained in the following manner: Obtain historical data; the historical data includes multiple influencing factors in previous construction processes; Obtain the carbon emission values in previous construction processes to obtain historical carbon emission values; Obtain a multiple regression model; Use multiple influencing factors in previous construction processes as the input of the multiple regression model, the historical carbon emission value as the label data, and the first carbon emission prediction value as the output, and train the multiple regression model to obtain the first carbon emission prediction model.

4. A method for predicting and correcting carbon emissions during construction according to claim 1, characterized in that, The first set value satisfies Formula 1: Formula 1; Among them, represents the first set value, T 总 represents the total construction duration, C 0 represents the carbon emission standard value of the entire construction process, t’ represents the monitoring duration from the first moment to the second moment, z represents the error tolerance threshold.

5. A method for predicting and correcting carbon emissions during construction according to claim 1, characterized in that, The weighted relationship is: Formula II; Among them, represents a weighted relationship, and its value is equal to the total value of the first error, represents the i weight of the error caused by the th i influence factor, represents the i error caused by the n th i influence factor, n .

6. The carbon emission prediction and correction method during construction according to claim 5, wherein, Use Formula 3 to perform non-linear correction on the weighted relationship to obtain a corrected total error expression; Formula III; Among them, represents the carbon emission prediction value after non-linear correction, represents the first carbon emission prediction value, represents the correction coefficient, represents the non-linear correction factor, represents the weighted relationship formula, represents the total correction error value expression; The correction parameter includes the correction coefficient and the non-linear correction factor.

7. A method for predicting and correcting carbon emissions during construction according to claim 6, characterized in that, Perform local optimization on the corrected total error expression to obtain the optimized weighted relationship; Local optimization includes: Obtain the actual carbon emission sub-values generated by each influencing factor during the period from the first moment to the second moment; Obtain the error sub-values brought by each influencing factor; Calculate the ratios of the error sub-values of multiple influencing factors to the actual carbon emission sub-values to obtain multiple first ratios; According to the multiple first ratios, re-allocate weights for multiple influencing factors to obtain optimized weights; Calculate the mean value of the multiple first ratios to obtain a first mean value; Take the impact factors corresponding to multiple first ratios greater than the first mean value as the local optimization objectives; Use the genetic algorithm to optimize the correction coefficient and the nonlinear correction factor of the local optimization objective to obtain the optimized correction coefficient and the optimized nonlinear correction factor; Substitute the optimized correction coefficient, the optimized nonlinear correction factor, and the optimized weight into Formula 2 to obtain the optimized weighted relationship formula.

8. A method for predicting and correcting carbon emissions during construction according to claim 1, characterized in that, According to the optimized weighted relationship formula, use the genetic algorithm to optimize the first model parameters and the correction parameters of the first carbon emission prediction model to obtain the optimized model parameters and the optimized correction parameters, including: Set the algorithm parameters, the fitness function, and the termination conditions; the algorithm parameters include the crossover probability and the mutation probability; Before the genetic algorithm optimization iteration, update the values of the crossover probability and the mutation probability according to the first total error value; During the genetic algorithm optimization iteration, make the crossover probability and the mutation probability decrease as the number of iterations increases; After the genetic algorithm reaches the termination conditions, optimize to obtain the historical optimal model parameters and the historical optimal correction parameters; Under the condition that the loss function satisfies gradient descent, adjust the historical optimal model parameters to obtain the optimized model parameters; Take the historical optimal correction parameters as the optimized correction parameters.

9. A carbon emission prediction and correction method during construction according to claim 8, characterized in that, Update the values of the crossover probability and the mutation probability according to the first total error value, including: Obtain the initial values of the crossover probability and the mutation probability to get the initial crossover probability and the initial mutation probability; Calculate the ratio of the first total error value to the actual carbon emission value to get the second ratio; Make the value of the crossover probability equal to the initial crossover probability plus the product of the initial crossover probability and the second ratio; Make the value of the mutation probability equal to the initial mutation probability plus the product of the initial mutation probability and the second ratio.

10. A carbon emission prediction and correction method during construction according to claim 8, characterized in that The genetic algorithm also has an evolutionary equation with a nonlinear penalty function as shown in Formula 4: Formula Four; Among them, represents the individual after the t +1 generation of the genetic algorithm, represents the individual of the current t generation, t represents the number of current iterations, and V are both adjustment factors for adjusting the evolutionary step size, represents the adaptive fitness function, represents the gradient of the adaptive fitness function, represents the random perturbation factor, represents the random perturbation term, G represents the individual of the genetic algorithm.

Images