Drill jumbo working hour prediction method and device, electronic equipment and storage medium

Abstract

The invention discloses a drill jumbo working hour prediction method and device, electronic equipment and a storage medium, and belongs to the technical field of working hour prediction, and the method comprises the steps: updating an original particle swarm position formula of an SABO algorithm according to Cubic chaotic mapping, and generating a first SABO algorithm; updating an initial iteration formula of the first SABO algorithm according to Piecewise chaotic mapping, and generating a second SABO algorithm; updating a position iteration formula of the second SABO algorithm according to a golden sine algorithm, and generating an optimal GSABO algorithm; performing iterative training on an initial neural network model based on the optimal GSABO algorithm and the index data to obtain an operation man-hour prediction model; and performing drill jumbo working hour prediction on the target operation based on the working hour prediction model. According to the method, the model parameters are determined by optimizing the SABO algorithm through the two chaotic mapping and golden sine algorithms, and the accuracy and efficiency of predicting the working hours of the drill jumbo are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of man-hour prediction, and in particular to a method, device, electronic equipment and storage medium for predicting the man-hour of a rock drilling rig operation. Background Art

[0002] With the continuous expansion of the mining depth and scale of underground mines and the continuous improvement of the automation of mechanical equipment, how to improve the utilization rate of rock drilling equipment and achieve efficient scheduling between different mining sites is an issue that every mining company needs to focus on. The prediction of the working hours of rock drilling trolleys is one of the important basic supports for the precise scheduling of rock drilling equipment. It is of great significance to better realize the real-time scheduling and precise control of equipment and provide data support for the construction of intelligent scheduling systems for underground mines.

[0003] In the prior art, the prediction of drilling rig operating hours usually involves building statistical models, machine learning models, deep learning models, etc., predicting multiple models through historical data, and then integrating the prediction results. Multiple models have uncertainty and waste of resources in the model training process, which increases the difficulty of model training and calculation, and makes the prediction results lack accuracy.

[0004] Therefore, there is an urgent need for a method to predict the operating hours of a rock drilling rig, which can optimize the model by optimizing the model parameters, simplify the model training and calculation process, and improve the model's accuracy and prediction efficiency. Summary of the invention

[0005] In view of this, it is necessary to provide a method, device, electronic device and storage medium for predicting the operating hours of a rock drilling rig, which can optimize the model by optimizing the model parameters, so as to simplify the model training and calculation process and improve the accuracy and prediction efficiency of the model.

[0006] In order to solve the above technical problems, on the one hand, the present invention provides a method for predicting the working hours of a rock drilling rig, comprising: The original particle swarm position formula of the SABO algorithm is updated according to the Cubic chaotic map to generate the first SABO algorithm; the initial iteration formula of the first SABO algorithm is updated according to the Piecewise chaotic map to generate the second SABO algorithm; the position iteration formula of the second SABO algorithm is updated according to the golden sine algorithm to generate the optimal GSABO algorithm; Determine the indicator data that affects the working hours of the operation; Iteratively training the initial neural network model based on the optimal GSABO algorithm and the indicator data to obtain a job working time prediction model; The operating time prediction model is used to predict the operating time of the drilling rig based on the acquired target index data of the target operation.

[0007] In a possible implementation, the original particle swarm position formula of the SABO algorithm is updated according to the Cubic chaotic map to generate the first SABO algorithm, including: Based on the initial particle swarm position formula of the SABO algorithm, the initial dimension position of the particles is determined according to the randomly generated initial dimension function; Based on the Cubic chaos mapping system, according to the initial dimensional function and the initial dimensional position, the second dimensional function of the particle is determined, and the second dimensional position of the particle is determined; According to the last dimension position and dimension function of the particle, the next dimension function of the particle is determined, and the next dimension position of the particle is determined to obtain the particle group position formula; The SABO algorithm is updated according to the particle swarm position formula to generate a first SABO algorithm.

[0008] In a possible implementation, the initial iteration formula of the first SABO algorithm is updated according to the Piecewise chaotic map to generate the second SABO algorithm, including: Based on the Piecewise chaotic mapping system, a segmented control factor is determined, and a segmented distribution of the segmented control factor is determined according to the segmented control factor; Determining an iterative function formula according to the piecewise chaotic mapping system and the piecewise distribution of the piecewise control factor; The iterative function formula is used to replace the initial random iterative function in the initial iterative formula to generate a second SABO algorithm.

[0009] In a possible implementation, the position iteration formula of the second SABO algorithm is updated according to the golden sine algorithm to generate the optimal GSABO algorithm, including: Based on the position iteration formula of the second SABO algorithm, the position of the current generation of particles is iteratively generated according to the position of the previous generation of particles; Based on the golden sine algorithm, the fitness values ​​of the previous generation particle position and the current generation particle position are calculated respectively according to the preset particle iteration position fitness function; The fitness value of the particle position is determined. When the fitness value does not change, the current generation particle position is recalculated according to the previous generation particle position based on the golden sine algorithm to generate an optimal iterative formula.

[0010] In a possible implementation, the initial neural network model is iteratively trained based on the optimal GSABO algorithm and the indicator data to obtain a job time prediction model, including: Based on the particle swarm position formula in the optimal GSABO algorithm, an initial L2 regularization coefficient, a learning rate, and a number of hidden units are randomly generated to form initial hyperparameters; Based on the convolutional neural network, the bidirectional long short-term memory network and the attention mechanism, an initial neural network model is constructed according to the initial hyperparameters; Based on the optimal iterative formula in the optimal GSABO algorithm, the hyperparameters are iterated to generate iterative hyperparameters, and the initial neural network model is trained according to the indicator data and the iterative hyperparameters in turn; The optimal iteration result is determined according to the preset target job fitness function, and the optimal hyperparameters are determined to obtain the job working time prediction model.

[0011] In one possible implementation, determining the indicator data that affects the working hours of the operation includes: Acquire information data of rock drilling rig operation, wherein the information data includes operation information data and corresponding operation hours; Based on the correlation analysis algorithm, the correlation between the operation information data and the operation hours is analyzed, and the information data with correlation higher than the first preset threshold is screened out to generate the first indicator data; Based on the multiple regression analysis algorithm, the regression coefficient between the operation information data and the corresponding operation hours in the first indicator data is calculated, the first indicator data whose regression coefficient is higher than the second preset threshold is screened out, and the indicator data affecting the operation hours is generated.

[0012] In one possible implementation, the optimal iteration result is determined according to a preset target job fitness function, and the optimal hyperparameter is determined to obtain a job man-hour prediction model, including: Determine the target job fitness function; Calculating the fitness value of the predicted working hours outputted in each iteration during the iterative training process and the corresponding working hours in the indicator data; According to the change of the fitness value and the maximum number of iterations, the optimal iteration result is determined, and the corresponding hyperparameter is the optimal hyperparameter, and the operation time prediction model is obtained.

[0013] In a second aspect, the present invention further provides a drilling rig operation time prediction device, comprising: The algorithm optimization module updates the original particle swarm position formula of the SABO algorithm according to the Cubic chaotic map to generate the first SABO algorithm; updates the initial iteration formula of the first SABO algorithm according to the Piecewise chaotic map to generate the second SABO algorithm; updates the position iteration formula of the second SABO algorithm according to the golden sine algorithm to generate the optimal GSABO algorithm; Determine the indicator data module and determine the indicator data that affects the operation hours; A model generation module, which iteratively trains the initial neural network model based on the optimal GSABO algorithm and the indicator data to obtain a job working time prediction model; The working time prediction module predicts the working time of the drilling rig based on the target index data of the acquired target operation based on the working time prediction model.

[0014] In a third aspect, the present invention further provides an electronic device comprising a memory and a processor, wherein the memory is used to store programs and data; the processor is coupled to the memory, and is used to execute the program stored in the memory, to implement the drilling rig operation time prediction method as described above, and / or to implement the drilling rig operation time prediction as described above.

[0015] In a fourth aspect, the present invention further provides a computer storage medium for storing computer-readable programs or instructions, which, when executed by a processor, can implement the drilling rig operation time prediction method as described above.

[0016] The beneficial effects of the present invention are as follows: first, the original particle swarm position formula and the initial iteration formula of the SABO algorithm are updated according to the Cubic chaotic mapping system and the Piecewise chaotic mapping, and then the position iteration formula is updated again according to the golden sine algorithm to generate the optimal GSABO algorithm; then, the initial neural network model is iteratively trained based on the optimal GSABO algorithm and the index data to obtain the operation time prediction model, and finally, the operation time of the drilling trolley is predicted according to the operation time prediction model and the index data of the target operation of the drilling trolley. The present invention optimizes the SABO algorithm by two different chaotic mappings, and further optimizes the SABO algorithm by the golden sine algorithm to obtain the optimal GSABO algorithm. The optimal GSABO algorithm has better chaotic ergodicity, enhances the randomness and diversity of particles, enhances the uniformity of particle distribution, and jumps out of the local optimal solution, so that the GSABO algorithm can quickly obtain the optimal hyperparameters during the model training process, and does not take up too much training and calculation time, and can optimize the model by optimizing the model parameters to simplify the model training and calculation process, while improving the accuracy and efficiency of the model. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0018] Figure 1A schematic flow chart of an embodiment of a method for predicting the working hours of a rock drilling rig provided by the present invention; Figure 2 For the present invention Figure 1 Schematic diagram of the process of the first embodiment of step S101; Figure 3 For the present invention Figure 1 A schematic flow chart of a second embodiment of step S101; Figure 4 For the present invention Figure 1 A schematic flow chart of a third embodiment of step S101; Figure 5 For the present invention Figure 1 A schematic diagram of a flow chart of an embodiment of step S102; Figure 6 For the present invention Figure 1 A schematic flow chart of an embodiment of step S103; Figure 7 For the present invention Figure 6 A schematic diagram of a flow chart of an embodiment of step S604; Figure 8 A schematic diagram of the structure of an embodiment of an electronic device for predicting the working hours of a rock drilling rig provided by the present invention; Fig. 9 A schematic diagram of the structure of an embodiment of a rock drilling rig operation time prediction storage medium provided by the present invention; Fig.10 A comparison chart of the predicted working hours and the true formula provided by the present invention; Fig.11 This is the change of fitness value during the iterative training process provided by the present invention. DETAILED DESCRIPTION

[0019] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0020] In the description of the embodiments of the present invention, unless otherwise specified, "multiple" means two or more than two. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" may mean: A exists alone, A and B exist at the same time, and B exists alone.

[0021] The descriptions of "first", "second", etc. involved in the embodiments of the present invention are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the technical features defined as "first" or "second" may explicitly or implicitly include at least one of the features.

[0022] Reference to an "embodiment" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiment may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0023] Before describing the embodiments, the following definitions are given for the relevant terms: Cubic chaotic mapping is a chaotic system based on cubic function, which generates chaotic sequences through nonlinear transformation of cubic function. Piecewise chaotic mapping is a piecewise defined chaotic system that divides the entire domain into multiple subintervals and applies different nonlinear transformations to each subinterval. The GSA (Golden Sine Algorithm) algorithm is the golden sine algorithm, which is an optimization algorithm based on the golden ratio (1.618). It simulates the law of the golden section, searches for the solution space by scanning the sine function within the unit circle, and narrows the search space through the golden ratio to approach the optimal solution.

[0024] The SABO algorithm uses the subtraction mean of multiple agents to update the positions of population members in the search space, and finds the best or near-best solution to the problem through search and optimization methods. The advantages of the SABO algorithm include strong optimization ability and fast convergence speed, which enables it to effectively solve complex optimization problems.

[0025] The present invention provides a method, device, electronic device and storage medium for predicting the operation time of a rock drilling rig, which are described below respectively.

[0026] Figure 1 A schematic flow chart of an embodiment of a method for predicting the working hours of a rock drilling rig provided by the present invention is shown in FIG. Figure 1 As shown in FIG. 1 , the method for predicting the working hours of a drilling rig includes: S101, updating the original particle swarm position formula of the SABO algorithm according to the Cubic chaotic map to generate a first SABO algorithm; updating the initial iteration formula of the first SABO algorithm according to the Piecewise chaotic map to generate a second SABO algorithm; updating the position iteration formula of the second SABO algorithm according to the golden sine algorithm to generate an optimal GSABO algorithm; S102, determining the index data that affects the operation hours; S103, iteratively training the initial neural network model based on the optimal GSABO algorithm and the indicator data to obtain a job working time prediction model; S104, predicting the operating hours of the drilling rig based on the acquired target indicator data of the target operation based on the operating hours prediction model.

[0027] It should be noted that the SABO algorithm first uses a random function to assign an initial position to each particle within the given upper and lower limit space, calculates the arithmetic mean of the initial position of each particle, and updates the particle position based on the difference between the initial position and the mean to iterate the particles.

[0028] It should be further explained that the Cubic chaotic mapping system is used to replace the generation of random functions to optimize the SABO algorithm, and the Piecewise chaotic mapping system is used to replace the generation of random iterative vector functions to further optimize the SABO algorithm. The golden sine algorithm is used to further optimize the position updates generated during the iteration process to obtain the optimal GSABO algorithm.

[0029] This embodiment updates the original particle swarm position formula and the initial iteration formula of the SABO algorithm according to the Cubic chaotic mapping system and the Piecewise chaotic mapping, and then updates the iteration formula again according to the golden sine algorithm to generate the optimal GSABO algorithm; iteratively trains the initial neural network model based on the optimal GSABO algorithm and the index data to obtain the operation time prediction model, and predicts the operation time of the drilling trolley according to the index data of the target operation of the drilling trolley based on the operation time prediction model. This embodiment optimizes the SABO algorithm by two different chaotic mappings, and further optimizes the SABO algorithm by the golden sine algorithm to obtain the optimal GSABO algorithm. The optimal GSABO algorithm has better chaotic ergodicity, enhances the randomness and diversity of particles, enhances the uniformity of particle distribution, and jumps out of the local optimal solution, so that the GSABO algorithm can quickly obtain the optimal hyperparameters during the model training process, and does not take up too much training and calculation time, and can optimize the model by optimizing the model parameters, simplifying the model training and calculation process, and also improving the accuracy and efficiency of the model.

[0030] This embodiment optimizes the SABO algorithm through two chaotic mapping systems and the golden sine algorithm, and uses the optimized SABO algorithm to obtain the optimal hyperparameters for network model training. The optimal hyperparameters can be obtained quickly without taking up too much training and computing time. The model can be optimized by optimizing the model parameters, which simplifies the model training and calculation process, and also improves the accuracy and efficiency of the model.

[0031] In some embodiments of the present invention, Figure 2 As shown, Figure 2 The present invention provides Figure 1 The flowchart of the first embodiment of step S101 in FIG. 1 is to update the original particle swarm position formula of the SABO algorithm according to the Cubic chaotic map to generate the first SABO algorithm, including: S201, based on the initial particle swarm position formula of the SABO algorithm, determining the initial dimension position of the particles according to the randomly generated initial dimension function; S202, based on the Cubic chaos mapping system, determining the second dimensional function of the particle according to the initial dimensional function and the initial dimensional position, and determining the second dimensional position of the particle; S203, determining the next dimensional function of the particle according to the previous dimensional position and dimensional function of the particle, and determining the next dimensional position of the particle, and obtaining the particle group position formula; S204, updating the SABO algorithm according to the particle swarm position formula to generate a first SABO algorithm.

[0032] It should be noted that according to the initial particle swarm position formula of the SABO algorithm, a random function is used to assign an initial position to each particle within the given upper and lower limit space. The initial particle swarm position has a significant impact on the final performance of the SABO algorithm. Therefore, the use of the Cubic chaotic mapping system to replace the randomly generated particle function can generate uniformly distributed initial particle positions throughout the entire space.

[0033] Specifically, the initial particle swarm position formula of the SABO algorithm is: , ,

[0034] in, is the particle number, is the particle dimension number, is the position of the particle, is the lower limit of the particle space, is the upper limit of the particle space, is a randomly generated particle function in the range [0,1], is the total number of particles, is the total number of particle dimensions; Use Cubic Chaos Mapping System to replace the randomly generated particle function , particle function The formula is: , ,

[0035] in, is the particle number, is the particle dimension number, is the position of the particle, is the particle function, is the chaos parameter. In this embodiment , is the total number of particles, is the total number of particle dimensions; According to the above formula, the dimensional positions of all particles can be generated in sequence according to the randomly generated initial dimensional function.

[0036] This embodiment optimizes the SABO algorithm by using a Cubic chaotic mapping system, which can generate uniformly distributed initial particle positions throughout the entire search space, which helps the algorithm to search in a global range and improves the accuracy of subsequent calculations.

[0037] In some embodiments of the present invention, Figure 3 As shown, Figure 3 The present invention provides Figure 1 The flowchart of the second embodiment of step S101 in FIG. 1 is to update the initial iteration formula of the first SABO algorithm according to the Piecewise chaotic map to generate a second SABO algorithm, including: S301, determining a segmentation control factor based on a Piecewise chaotic mapping system, and determining a segmentation distribution of the segmentation control factor according to the segmentation control factor; S302, determining an iterative function formula according to the piecewise chaotic mapping system and the piecewise distribution of the piecewise control factor; S303: Use the iterative function formula to replace the initial random iterative function in the initial iterative formula to generate a second SABO algorithm.

[0038] It should be noted that in each iterative step of the SABO algorithm, the arithmetic mean of the positions of all particles is calculated, and the position of each particle is updated by combining the difference between each particle and the mean and the randomly generated iterative vector function. In this process, the Piecewise chaotic mapping system is added, and a suitable piecewise distribution of the piecewise control factor is designed. The Piecewise chaotic mapping system is used to replace the randomly generated iterative vector function, which can make the iteratively generated particle position distribution more uniform, so that the SABO algorithm increases the diversity of iterative particles when making the average difference.

[0039] Specifically, the initial iteration formula of the SABO algorithm is: , , in, is the particle number, is the iterative position of the particle, is the particle's previous position, is a randomly generated iterative vector function in the range [1,2], To control the parameters, is the total number of particles; Use Piecewise Chaotic Mapping System to Replace Randomly Generated Iterated Vector Function , iterated vector function The formula is: , in, is the particle number, is the particle iteration vector function, is the segment control factor, The value range of is [0, 0.5]. ; According to the above formula, an iterative function can be generated according to the iterative function formula to replace the randomly generated iterative function, and the iterative position of the particle can be calculated according to the iterative vector function.

[0040] This embodiment optimizes the iteration of the SABO algorithm by using a Piecewise chaotic mapping system, which can make the position distribution of iteratively generated particles more uniform, increase the diversity of iterative particles, and improve the accuracy of subsequent calculations.

[0041] In some embodiments of the present invention, Figure 4 As shown, Figure 4 The present invention provides Figure 1 The flowchart of the third embodiment of step S101 in FIG. 1 is to update the position iteration formula of the second SABO algorithm according to the golden sine algorithm to generate the optimal GSABO algorithm, including: S401, based on the position iteration formula of the second SABO algorithm, iteratively generate the position of the current generation of particles according to the position of the previous generation of particles; S402, respectively calculating the fitness values ​​of the particle positions of the previous generation and the particle positions of the current generation according to a preset particle position fitness function; S403, determining the fitness value of the particle position, and when the fitness value does not change, recalculating the iterative position of the current generation of particles according to the iterative position of the previous generation of particles based on the golden sine algorithm to obtain the optimal iterative formula.

[0042] It should be noted that the SABO algorithm uses the positions of all particles as the subtracted average to iteratively update the positions. When the particles are not in a good initial position, it is very easy to fall into a local optimal solution. The golden sine algorithm is introduced to re-update the iterative position of particles with a bad initial position using the golden sine algorithm to obtain a new optimal iterative position.

[0043] Specifically, the fitness value of the previous generation particle position and the fitness value of the current generation particle position are calculated respectively. When the fitness value changes as expected, the current generation particle position is the optimal iteration position. If the fitness value does not change, the golden sine algorithm is used to recalculate the iteration position of the particle. The position iteration formula for calculating the iteration position of the particle using the golden sine algorithm is: , , in, is the iterative position of the particle, is the particle number, For the The iterative position of the particle in the previous generation, is the optimal iterative position of all particles in the previous generation, and are random numbers in the range of [0,2π] and [0,π] respectively, and is the golden ratio coefficient; According to the above formula, the optimal iteration position can be recalculated according to the previous iteration position of the particle.

[0044] This embodiment optimizes the iteration of the SABO algorithm by using the golden sine algorithm, so that the SABO algorithm jumps out of the local optimal solution during the iteration process, thereby improving the accuracy of subsequent calculations.

[0045] In some embodiments of the present invention, Figure 5 As shown, Figure 5 The present invention provides Figure 1 The flowchart of an embodiment of step S102 in the embodiment includes: S501, obtaining information data of the rock drilling rig operation, the information data including operation information data and corresponding operation hours; S502: Analyze the correlation between the operation information data and the operation hours based on a correlation analysis algorithm, select information data with a correlation higher than a first preset threshold, and generate first indicator data; S503, based on the multivariate regression analysis algorithm, calculate the regression coefficient between the operation information data in the first indicator data and the corresponding operation hours, screen out the first indicator data whose regression coefficient is higher than the second preset threshold, and generate indicator data that affects the operation hours.

[0046] It should be noted that when obtaining information data of the drilling rig operation, different working faces are selected respectively, and before the drilling rig on the same working face performs drilling operations, typical rock blocks of the working face are selected, and samples are taken according to the core pattern size that meets the production standards; the obtained rock blocks are processed into cores for indoor testing, and the saturated uniaxial compressive strength is measured, which represents the rock strength index of the working face; during the drilling operation of the drilling rig on the working face, the real-time drilling influencing factor indicators such as drill rod rotation speed, thrust pressure, rotation pressure, impact pressure, drilling power, etc. are measured at stable operating time points at a certain interval during the drilling of a single hole, and the corresponding borehole diameter, borehole depth and real-time drilling speed are recorded; the rock strength index, drilling influencing factor index, borehole diameter, borehole depth and real-time drilling speed are used as operation information data; the real-time drilling speed is converted into drilling time per unit depth, which is recorded as operation hours, that is, the corresponding operation hours.

[0047] It should be further explained that the rock strength index collected on the same working face is used to construct a data set in one-to-one correspondence with the real-time drilling influencing factor index data and working hours of the working face.

[0048] Specifically, a Pearson correlation analysis is performed on the operation information data, and the Pearson correlation of each influencing factor indicator is calculated. The influencing factor indicators with correlation less than the preset threshold are eliminated, and then the regression equation, Durbin-Watson test value, variance inflation factor, etc. of the remaining influencing factor indicators are analyzed according to the multiple regression analysis method. According to the respective preset thresholds, the index information data that is closely related to the influencing factors of the drilling rig operation hours are finally screened out.

[0049] This embodiment performs correlation analysis and multivariate regression analysis on the information data of the rock drilling rig operation to screen out indicator data that has a high impact on the operating hours of the rock drilling rig. By training the model with these influencing factor indicator data, the accuracy of the model prediction can be improved, and the efficiency of the prediction is also improved.

[0050] In some embodiments of the present invention, Figure 6 As shown, Figure 6 The present invention provides Figure 1 The flowchart of an embodiment of step S103 in the embodiment includes: S601, randomly generating an initial L2 regularization coefficient, a learning rate, and a number of hidden units based on the particle swarm position formula in the optimal GSABO algorithm to form initial hyperparameters; S602, constructing an initial neural network model according to initial hyperparameters based on a convolutional neural network, a bidirectional long short-term memory network, and an attention mechanism; S603, based on the optimal iteration formula in the optimal GSABO algorithm, iterate the hyperparameters to generate iterative hyperparameters, and train the initial neural network model according to the indicator data and the iterative hyperparameters in turn; S604: Determine the optimal iteration result according to the preset target job fitness function, and determine the optimal hyperparameters to obtain the job man-hour prediction model.

[0051] It should be noted that the initial GSABO-CNN-BiLSTM-SE Attention is constructed according to the convolutional neural network, the bidirectional long short-term memory network and the attention mechanism combined with the optimal GSABO algorithm, and the data set of operation hours is obtained. The data set is randomly shuffled, 80% of the data set is used to train the GSABO-CNN-BiLSTM-SE Attention model, and the remaining 20% ​​of the data set is used as the test set to test the accuracy of the model. In this embodiment, the size of the training data set is 1832, the size of the test data set is 458, and the determination coefficient R of the predicted value is 1. 2 The mean percentage absolute error (MAPE) is used as the model performance evaluation index to iteratively generate the final operation time prediction model, and the determination coefficient R 2 It reaches 0.9839, which is greater than 0.95, the MAPE is 6.64%, and the overall accuracy is 93.36%, which is greater than 90%, meeting the actual needs of the project. Fig.10 As shown, Fig.10 A comparison chart of the predicted working hours and the true formula provided by the present invention.

[0052] Specifically, the initial L2 regularization coefficient, learning rate and number of hidden units are generated based on the optimal GSABO algorithm. These three hyperparameters have a greater impact on the model performance and are closely related to the model fitting degree and convergence speed. The main body of the CNN-BiLSTM-SE Attention model is constructed based on these three initial hyperparameters. The feature weights are automatically adjusted according to the importance of the input data, and features are extracted and the influence of the feature state on the predicted value is analyzed. The CNN-BiLSTM-SE Attention model has many parameters and a long training time. The optimal GSABO algorithm is used to iteratively optimize the parameters, and the output of each iterative model is calculated at the same time. The optimal L2 regularization coefficient, learning rate and number of hidden units are selected to construct the optimal drilling rig operation time prediction model GSABO-CNN-BiLSTM-SE Attention model. The optimal GSABO has excellent optimization ability and does not occupy too many computing resources, and can also improve the model accuracy.

[0053] In this embodiment, the operation time prediction model is composed of multiple key components, including an input layer, a CNN layer (further subdivided into a convolutional layer and a pooling layer), a BiLSTM layer (consisting of a forward LSTM layer and a backward LSTM layer), a SE Attention layer, and an output layer. The GSABO algorithm optimization parameters are set as follows: population size 20, maximum number of iterations 100, number of optimization parameters 3, and upper and lower limits of the optimization parameters are [0.01, 0.1], [10, 30], and [0.001, 0.0001], respectively. After 61 optimization iterations, the optimal initial learning rate, number of hidden units, and L2 regularization parameter are 0.0055436, 10, and 0.0001006, respectively. The corresponding fitness curve is shown in the figure. Fig.11 As shown, the fitness is a curve. As can be seen from the figure, when the fitness value reaches 4.63751, it no longer changes. At this time, the number of iterations is 61, and the optimal hyperparameters are obtained when the iterations are 61 times.

[0054] This embodiment uses the optimal GSABO algorithm to iteratively train the neural network model according to the indicator data. The optimal GSABO algorithm has excellent optimization ability and does not occupy too many computing resources. The optimal hyperparameters of the model are screened out through the optimal GSABO algorithm, which greatly improves the model accuracy and efficiency.

[0055] In some embodiments of the present invention, Figure 7 As shown, Figure 7 The present invention provides Figure 6 The flowchart of step S604 in an embodiment includes: S701, determining the fitness function of the target operation; S702, calculating the fitness value of the predicted working hours outputted by each generation in the iterative training process and the corresponding working hours in the indicator data; S703. According to the change of the fitness value and the maximum number of iterations, the optimal iteration result is determined, and the corresponding hyperparameter is the optimal hyperparameter, so as to obtain a job working time prediction model.

[0056] It should be noted that this embodiment uses the root mean square error (RMSE) as the fitness function of the target job, and calculates the fitness value based on the output of each iterative training model and the corresponding job hours in the training data. When the fitness value has not changed after multiple iterations, or the fitness value has reached the preset minimum value, or the number of iterations has reached the maximum iteration limit, the corresponding model parameters are the hyperparameters of the job hour prediction model, thereby obtaining the optimal job hour prediction model.

[0057] This embodiment uses the root mean square error (RMSE) as the fitness function of the target job, and determines the hyperparameters of the model according to the fitness value, so that the model can accurately and efficiently predict the target job.

[0058] In order to better implement the method for predicting the working hours of a rock drilling rig in the embodiment of the present invention, based on the method for predicting the working hours of a rock drilling rig, correspondingly, Figure 8 As shown, the embodiment of the present invention further provides a drilling rig operation time prediction device 800, comprising: Determine indicator data module 801, determine indicator data that affects operation hours; The algorithm optimization module 802 updates the original particle swarm position formula of the SABO algorithm according to the Cubic chaotic map to generate a first SABO algorithm; updates the initial iteration formula of the first SABO algorithm according to the Piecewise chaotic map to generate a second SABO algorithm; updates the position iteration formula of the second SABO algorithm according to the GSA algorithm to generate an optimal GSABO algorithm; The model generation module 803 iteratively trains the initial neural network model based on the optimal GSABO algorithm and the indicator data to obtain a job time prediction model; The working time prediction module 804 predicts the working time of the drilling rig based on the target indicator data of the acquired target operation based on the working time prediction model.

[0059] The drilling rig operation time prediction device 800 provided in the above embodiment can implement the technical solution described in the above embodiment of the drilling rig operation time prediction method. The specific implementation principles of the above modules or units can refer to the corresponding contents in the above embodiment of the drilling rig operation time prediction method, which will not be repeated here.

[0060] In the embodiment of the present invention, the drilling rig operation time prediction device can be an independent server, or a server network or server cluster composed of servers. For example, the drilling rig operation time prediction device described in the embodiment of the present invention includes but is not limited to a computer, a network host, a single network server, a plurality of network server sets or a cloud server composed of multiple servers. The cloud server is composed of a large number of computers or network servers based on cloud computing.

[0061] The present invention also provides a corresponding electronic device for predicting the working hours of a rock drilling rig. Fig. 9 As shown, Fig. 9 The structural block diagram of an embodiment of the drilling rig operation time prediction device provided by the present invention. The drilling rig operation time prediction device 900 can be a computing device such as a mobile terminal, a desktop computer, a notebook, a palm computer, and a server. The drilling rig operation time prediction device 900 includes a processor 901 and a memory 902, wherein the memory 902 stores a drilling rig operation time prediction program 903.

[0062] In some embodiments, the memory 902 may be an internal storage unit of a computer device, such as a hard disk or memory of the computer device. In other embodiments, the memory 902 may also be an external storage device of the computer device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the computer device. Furthermore, the memory 902 may also include both an internal storage unit of the computer device and an external storage device. The memory 902 is used to store application software and various types of data installed in the computer device, such as program codes for installing the computer device. The memory 902 may also be used to temporarily store data that has been output or is to be output. In one embodiment, the drilling rig operation time prediction program 903 may be executed by the processor 901, thereby realizing the drilling rig operation time prediction method, device, electronic device and storage device of each embodiment of the present invention.

[0063] In some embodiments, the processor 901 may be a central processing unit (CPU), a microprocessor or other data processing chip, used to run program codes or process data stored in the memory 902, such as executing a drilling rig operation time prediction program.

[0064] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods 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. Any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0065] The rock drilling rig operation time prediction method, device, electronic device and storage device provided by the present invention are introduced in detail above. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A method for predicting the working hours of a rock drilling rig, characterized in that: include: The original particle swarm position formula of the SABO algorithm is updated according to the Cubic chaotic map to generate the first SABO algorithm; the initial iteration formula of the first SABO algorithm is updated according to the Piecewise chaotic map to generate the second SABO algorithm; the position iteration formula of the second SABO algorithm is updated according to the golden sine algorithm to generate the optimal GSABO algorithm; Determine the indicator data that affects the working hours of the operation; Iteratively training the initial neural network model based on the optimal GSABO algorithm and the indicator data to obtain a job time prediction model; The operating time prediction model is used to predict the operating time of the drilling rig based on the acquired target index data of the target operation.

2. The method for predicting the working hours of a drilling rig according to claim 1, characterized in that: The original particle swarm position formula of the SABO algorithm is updated according to the Cubic chaotic map to generate the first SABO algorithm, including: Based on the initial particle swarm position formula of the SABO algorithm, the initial dimension position of the particles is determined according to the randomly generated initial dimension function; Based on the Cubic chaos mapping system, according to the initial dimensional function and the initial dimensional position, the second dimensional function of the particle is determined, and the second dimensional position of the particle is determined; According to the last dimension position and dimension function of the particle, the next dimension function of the particle is determined, and the next dimension position of the particle is determined to obtain the particle group position formula; The SABO algorithm is updated according to the particle swarm position formula to generate a first SABO algorithm.

3. The method for predicting the working hours of a drilling rig according to claim 2, characterized in that: The initial iteration formula of the first SABO algorithm is updated according to the Piecewise chaotic map to generate a second SABO algorithm, including: Based on the Piecewise chaotic mapping system, a segmented control factor is determined, and a segmented distribution of the segmented control factor is determined according to the segmented control factor; Determining an iterative function formula according to the piecewise chaotic mapping system and the piecewise distribution of the piecewise control factor; The iterative function formula is used to replace the initial random iterative function in the initial iterative formula to generate a second SABO algorithm.

4. The method for predicting the working hours of a drilling rig according to claim 3, characterized in that: The position iteration formula of the second SABO algorithm is updated according to the golden sine algorithm to generate the optimal GSABO algorithm, including: Based on the position iteration formula of the second SABO algorithm, the position of the current generation of particles is iteratively generated according to the position of the previous generation of particles; Calculating the fitness values ​​of the previous generation particle position and the current generation particle position respectively according to a preset particle iteration position fitness function; The fitness value of the particle position is determined. When the fitness value does not change, the current generation particle position is recalculated according to the previous generation particle position based on the golden sine algorithm to generate an optimal iterative formula.

5. The method for predicting the working hours of a drilling rig according to claim 1, characterized in that: Determine the indicators that affect the working hours of the operation, including: Acquire information data of rock drilling rig operation, wherein the information data includes operation information data and corresponding operation hours; Based on the correlation analysis method, the correlation between the operation information data and the operation hours is analyzed, and the information data with a correlation higher than a first preset threshold is screened out to generate first indicator data; Based on the multivariate regression analysis method, the regression coefficient between the operation information data and the corresponding operation hours in the first indicator data is calculated, the first indicator data whose regression coefficient is higher than the second preset threshold is screened out, and the indicator data affecting the operation hours is generated.

6. The method for predicting the working hours of a drilling rig according to claim 4, characterized in that: The initial neural network model is iteratively trained based on the optimal GSABO algorithm and the indicator data to obtain a job time prediction model, including: Based on the particle swarm position formula in the optimal GSABO algorithm, an initial L2 regularization coefficient, a learning rate, and a number of hidden units are randomly generated to form initial hyperparameters; Based on the convolutional neural network, the bidirectional long short-term memory network and the attention mechanism, an initial neural network model is constructed according to the initial hyperparameters; Based on the optimal iterative formula in the optimal GSABO algorithm, the hyperparameters are iterated to generate iterative hyperparameters, and the initial neural network model is trained according to the indicator data and the iterative hyperparameters in turn; The optimal iteration result is determined according to the preset target job fitness function, and the optimal hyperparameters are determined to obtain the job working time prediction model.

7. The method for predicting the working hours of a drilling rig according to claim 6, characterized in that: The optimal iteration result is determined according to the preset target job fitness function, and the optimal hyperparameters are determined to obtain the job man-hour prediction model, including: Determine the target job fitness function; Calculating the fitness value of the predicted working hours outputted in each iteration during the iterative training process and the corresponding working hours in the indicator data; According to the change of the fitness value and the maximum number of iterations, the optimal iteration result is determined, and the corresponding hyperparameter is the optimal hyperparameter, and the operation time prediction model is obtained.

8. A drilling rig operation time prediction device, characterized in that: include: The algorithm optimization module updates the original particle swarm position formula of the SABO algorithm according to the Cubic chaotic map to generate the first SABO algorithm; updates the initial iteration formula of the first SABO algorithm according to the Piecewise chaotic map to generate the second SABO algorithm; updates the position iteration formula of the second SABO algorithm according to the golden sine algorithm to generate the optimal GSABO algorithm; Determine the indicator data module and determine the indicator data that affects the operation hours; A model generation module, which iteratively trains the initial neural network model based on the optimal GSABO algorithm and the indicator data to obtain a job working time prediction model; The working time prediction module predicts the working time of the drilling rig based on the target index data of the acquired target operation based on the working time prediction model.

9. An electronic device for predicting the working hours of a rock drilling rig, characterized in that: It comprises a processor and a memory, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the steps of the method for predicting the working time of a drilling rig as claimed in any one of claims 1 to 7 are implemented.

10. A storage medium, characterized in that: The storage medium stores computer program instructions, and when the computer program instructions are executed by a computer, the computer executes the method for predicting the working hours of a drilling rig according to any one of claims 1 to 7.

Images