Intelligent recommendation method for mine support scheme environmental analogy database

Abstract

The present application belongs to the field of data processing and mine support technology, and specifically relates to a mine support scheme environment analogy database intelligent recommendation method, which comprises: obtaining multi-source time series data of a region to be supported and performing benchmarking preprocessing; constructing a dynamic mining disturbance response index according to the microseismic event energy and the hypocenter distance, in combination with the geological medium attenuation characteristics; based on the rock rheological damage theory, introducing a forgetting mechanism to perform time-varying weighted accumulation on the dynamic mining disturbance response index within a historical time window, and calculating a rock mass rheological damage cumulative factor; constructing a matching model containing a damage correction dimension, calculating a comprehensive matching degree score according to the static geological feature difference and the damage cumulative factor difference between the region to be supported and historical cases, and recommending support parameters. The present application solves the technical problem that single-time monitoring cannot reveal historical cumulative hidden dangers, and can effectively prevent the common lag-type roof fall accidents in deep mining.

Description

Technical Field

[0001] This invention relates to the fields of data processing and mine support technology, and in particular to an intelligent recommendation method for mine support schemes based on environmental analogy databases. Background Technology

[0002] In deep metal mining operations, a scientifically sound and reasonable design of roadways and stopes support is a core element in ensuring the safety of personnel and equipment. Traditional support design models typically rely on expert systems based on case-based reasoning. This involves collecting geological parameters of the current working conditions, searching for the most similar successful cases in a historical database, and then reusing their support parameters such as anchor spacing and shotcrete thickness.

[0003] However, existing data recommendation technologies have revealed significant limitations in practical applications. First, current technologies primarily rely on static geological indicators for matching, neglecting the dynamic disturbances caused by frequent blasting and tunneling in deep mining. This leads to two areas with identical static indicators but drastically different dynamic stress environments being incorrectly classified as similar. In deep mining environments, frequent dynamic stress significantly alters the stability of the surrounding rock, and static parameters cannot fully reflect this change. Furthermore, existing methods typically use simple Euclidean distance or cosine similarity to calculate differences between static features when matching cases, failing to specifically model and adjust weights for the crucial dimension of rock mass damage state.

[0004] Secondly, rock materials themselves possess rheological and memory properties. Existing recommendation algorithms often only focus on the monitoring values ​​at the current moment, failing to effectively assess the cumulative destructive effects over a period of time. This situation leads to the recommendation system underestimating the potential risks of areas that appear intact on the surface but have suffered severe fatigue damage due to historical disturbances, thus providing support recommendations with insufficient support resistance and creating potential safety hazards. Summary of the Invention

[0005] To address the specific technical problems in existing technologies, such as low matching degree and insufficient safety of support scheme recommendations due to neglecting the dynamic disturbance effects of frequent blasting during deep mining and the cumulative damage history caused by the rheological properties of rock mass, this invention provides an intelligent recommendation method for mine support schemes based on an environmental analogy database, including:

[0006] Acquire multi-source time-series data of the area to be supported and perform benchmarking preprocessing. The multi-source time-series data includes microseismic monitoring data, static geological data, and historical case database data.

[0007] Based on the energy value and source distance of the microseismic events in the microseismic monitoring data, and combined with the attenuation characteristics of the geological medium, a dynamic mining disturbance response index reflecting the instantaneous impact intensity of the current mining activity on the area to be supported is calculated.

[0008] Based on the theory of rock rheological damage, a forgetting mechanism is introduced to perform time-varying weighted accumulation of the dynamic mining disturbance response index within the historical time window, and to calculate the rock mass rheological damage accumulation factor that reflects the historical fatigue state of the rock mass.

[0009] A similarity matching model including a damage correction dimension is constructed. Based on the differences in static geological characteristics between the area to be supported and historical cases in the historical case database, as well as the differences in the cumulative factor of rock mass rheological damage, a comprehensive matching score is calculated. Target support parameters are then selected and recommended from the historical case database based on the comprehensive matching score.

[0010] This invention, by adopting the above-mentioned technical solution, constructs a dynamic and static combined support recommendation system. It not only considers static geological characteristics but also incorporates dynamic mining disturbances and historical cumulative damage of the rock mass into the evaluation system. It can accurately identify high-risk rock masses that are surface intact but internally fatigued, thereby recommending more targeted and safer redundant support parameters. This effectively prevents delayed roof collapse accidents commonly seen in deep mining and transforms historical microseismic monitoring data into damage state indicators that can be directly used for matching decisions, improving the targeting and safety redundancy of the recommended support schemes.

[0011] In one embodiment, the microseismic monitoring data includes the time of occurrence of the microseismic event, the coordinates of the seismic source, and the energy released; the static geological data includes rock quality indicators, uniaxial compressive strength of the rock, and joint orientation information; and the historical case database data includes geological environment descriptions of historical cases, implemented support parameters, and maintenance records.

[0012] By clarifying the specific composition of various types of data, a foundation was laid for subsequent data processing and feature extraction, ensuring the integrity and traceability of multi-source data.

[0013] In one embodiment, the calculation of the dynamic mining disturbance response index, which reflects the instantaneous impact intensity of the current mining activity on the area to be supported, is specifically expressed as follows:

[0014]

[0015] In the formula, Indicates time The dynamic mining disturbance response index. This indicates the total number of microseismic events detected within the current time window. Indicates the first The energy value released by this event. Indicates the first The Euclidean distance between the epicenter coordinates of the event and the center point of the area to be supported. Indicates the damping coefficient of the geological medium. Represents the power exponent of geometric decay of energy. This represents the energy normalization coefficient.

[0016] The dynamic mining disturbance response index constructed using this expression can comprehensively assess the impact of energy magnitude, source distance, and geological medium attenuation on impact intensity. It uses a logarithmic function to smooth the numerical fluctuations of extreme energy events and scientifically reflects the degree of instantaneous dynamic disturbance of the rock mass caused by mining activities.

[0017] In one embodiment, the damping coefficient of the geological medium is derived in reverse from the rock integrity coefficient, and its value is inversely proportional to the rock quality index; the energy geometrical decay power exponent is set according to the wave velocity decay law of the geological medium.

[0018] In one embodiment, the calculation of the rock mass rheological damage accumulation factor, which reflects the historical fatigue state of the rock mass, is specifically expressed as follows:

[0019]

[0020] In the formula, Indicates time Rock mass rheological damage accumulation factor Indicates the length of the effective traceability time window. Representing historical moments The dynamic mining disturbance response index. Indicates the rock mass self-healing rate coefficient. This represents the rock mass tolerance constant.

[0021] By introducing a forgetting mechanism for cumulative calculation, the fatigue evolution process of rock mass under repeated impacts is reflected, transforming discrete instantaneous disturbances into continuous rock mass damage states, thus solving the technical problem that single-moment monitoring cannot reveal historical accumulated hidden dangers.

[0022] In one embodiment, the rock mass self-healing rate coefficient is determined according to the rock type, with a smaller value for brittle rocks to characterize strong memory and a larger value for soft rocks to characterize weak memory; the rock mass tolerance capacity constant is determined by laboratory rock fatigue tests.

[0023] By setting parameters differently according to rock type, the damage accumulation model can be adapted to different geological conditions, improving the universality and accuracy of the method.

[0024] In one embodiment, calculating the overall matching score includes:

[0025]

[0026] In the formula, This indicates the current area awaiting support and the first case in the historical case database. The overall matching score of each case This represents the total number of static geological characteristic indicators. Indicates the first The weight of each static geological characteristic index, Indicates the current area to be supported. A static geological characteristic index, Indicates the first The first historical case A static geological characteristic index, This represents the cumulative rheological damage factor of the rock mass in the area to be supported at the current moment. Indicates the first The cumulative factor of rock mass rheological damage during support implementation in a historical case. This represents the penalty weight coefficient for the damage dimension. This represents the numerical stability constant.

[0027] In one embodiment, the weights of the static geological feature indicators are determined by principal component analysis; the penalty weight coefficient of the damage dimension is set to a constant greater than 1.

[0028] In one embodiment, the benchmarking preprocessing includes:

[0029] The energy data in the microseismic monitoring data are logarithmically processed to eliminate differences in magnitude;

[0030] The unstructured geological descriptions in the static geological data are mapped to numerical values ​​using an expert scoring method;

[0031] A unified time sampling window is set, and the continuously collected data is discretized and aligned according to the time sampling window.

[0032] By using systematic benchmarking preprocessing, the differences in dimensions and formats among multi-source heterogeneous data are eliminated, improving data consistency and comparability, and providing a high-quality input foundation for subsequent complex calculations.

[0033] In one embodiment, the recommended target support parameters include: selecting a preset number of historical cases with the highest comprehensive matching score, and calculating a weighted average of the support parameters of the historical cases to obtain the final recommended support parameters.

[0034] By weighted averaging the support parameters of multiple highly matched cases, the randomness of a single case is avoided, thus improving the stability and reliability of the recommendation results.

[0035] The technical solution of the present invention has the following beneficial technical effects:

[0036] This invention constructs a dynamic and static support recommendation system through the above-mentioned technical solution. It not only considers static geological characteristics, but also incorporates dynamic mining disturbances and historical cumulative damage of the rock mass into the evaluation system. It can accurately identify high-risk rock masses that are intact on the surface but fatigued internally, thereby recommending more targeted and safer redundant support parameters. This effectively prevents delayed roof collapse accidents that are common in deep mining, and achieves the organic unity of in-depth mining of historical data value and engineering safety assurance.

[0037] Furthermore, by introducing a forgetting mechanism to perform time-varying weighted accumulation of the dynamic mining disturbance response index within the historical time window, discrete instantaneous disturbances are transformed into continuous rock mass damage states. This solves the technical problem that single-moment monitoring cannot reveal historically accumulated hidden dangers, and significantly improves the safety and adaptability of deep mining support design. Attached Figure Description

[0038] Figure 1 This is a flowchart of an intelligent recommendation method for mine support schemes based on an environmental analogy database, according to an embodiment of the present invention.

[0039] Figure 2 This is a schematic diagram of the time-varying evolution process of rock mass damage under mining disturbance according to an embodiment of the present invention;

[0040] Figure 3 This is a comparison chart of the accuracy of recommended support schemes under complex working conditions according to embodiments of the present invention. Detailed Implementation

[0041] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0042] Figure 1 This is a flowchart of a method for intelligent recommendation of mine support schemes using an environmental analogy database, according to an embodiment of the present invention. Figure 1 As shown, the intelligent recommendation method for mine support schemes based on environmental analogy databases includes steps S101 to S104, which are described in detail below:

[0043] S101. Acquire multi-source time-series data of the area to be supported and perform benchmarking preprocessing. The multi-source time-series data includes microseismic monitoring data, static geological data, and historical case database data.

[0044] In this embodiment, data acquisition is a fundamental step in the intelligent recommendation method. First, microseismic monitoring data is collected in real time through a microseismic monitoring system deployed underground in the mine. For example, a Geophonearray sensor array is used. The microseismic monitoring data specifically includes the occurrence time of microseismic events, the three-dimensional coordinates of the seismic source, and the magnitude of the released energy, measured in joules. Second, static geological data of the area to be supported is obtained through geological drilling core recordings and face geological logging. This static geological data includes rock quality indicators, uniaxial compressive strength of the rock, and joint orientation information. Finally, historical case database data is extracted from the company's past engineering database. This database includes geological environment descriptions of historical cases, implemented support parameters, and subsequent maintenance records. For example, the support parameters include anchor bolt length and anchor bolt spacing.

[0045] After acquiring multi-source time-series data, it is necessary to perform benchmarking preprocessing. Considering the large span of energy data in microseismic monitoring data, the energy data is logarithmically converted to eliminate magnitude differences. Unstructured geological descriptions in static geological data are mapped to values ​​in the [0, 1] interval using an expert scoring method; for example, the description of relatively fractured rock mass is mapped to 0.3. A unified time sampling window is set, exemplarily on a daily basis, and continuously acquired data is discretized and aligned according to the time sampling window for subsequent time-series analysis.

[0046] By filtering and standardizing multi-source time series data as described above, a more accurate and unified dataset can be obtained, eliminating the dimensional barriers between different types of data in multi-source time series data, and providing a high-quality input foundation for subsequent complex calculations.

[0047] S102. Based on the energy value and source distance of microseismic events in the microseismic monitoring data, and combined with the attenuation characteristics of the geological medium, calculate the dynamic mining disturbance response index, which reflects the instantaneous impact intensity of the current mining activity on the area to be supported.

[0048] In this embodiment, since mining activities such as blasting and tunneling can cause rock mass fracturing and generate microseismic events, this invention captures these microseismic events through a microseismic monitoring system. However, since simply counting the number of microseismic events cannot accurately reflect the impact intensity, it is necessary to comprehensively consider the energy released by the microseismic events and the distance to the seismic source. The specific operation method in this embodiment is as follows:

[0049] First, determine the center coordinates of the area to be supported. For the current time window... For each microseismic event occurring within the area, calculate the instantaneous impact intensity of that event on the area to be supported. The specific calculation expression is as follows:

[0050]

[0051] In the formula, Indicates time The dynamic mining disturbance response index is a dimensionless value. This represents the total number of microseismic events detected within the current time window; Indicates the first The energy released by this event is measured in joules. Indicates the first The Euclidean distance between the epicenter of this event and the center of the area to be supported is expressed in meters. This represents the damping coefficient of the geological medium, with units of 1000g. The damping coefficient of the geological medium is expressed by the formula By reverse deduction, we can obtain that, The constant is the rock quality index. A lower value indicates a more fractured rock mass and a lower geological damping coefficient. The larger; This represents the power exponent of geometrical energy decay, which is set based on the measured wave velocity decay law in the mine, and is preferably set to 2. This represents the energy normalization coefficient, used to prevent numerical overflow.

[0052] For example, the following parameters are set for calculation demonstration:

[0053] Assuming energy normalization coefficient for Damping coefficient of geological medium The value is 0.02, representing the power exponent of geometric decay. The value is 2.

[0054] Suppose that at the current moment... The energy released by this micro-seismic event 10 7 Joules, the distance of the event from the center of the area to be supported. It is 50 meters.

[0055] First, calculate the distance attenuation term in the denominator: Then, the instantaneous impact contribution of the microseismic event to the supported area is calculated: Finally, the dynamic mining disturbance response index is calculated: .

[0056] If the microseismic event is farther from the center of the area to be supported, for example, when the distance is... When the distance is 500 meters, the denominator becomes very large, causing the effective impact contribution of the microseismic event to the supported area to decrease sharply, and the calculated dynamic mining disturbance response index... It will also be very small, possibly close to 0.

[0057] As can be seen from the above process, by constructing a dynamic mining disturbance response index, the actual impact intensity of each vibration event on the rock mass can be scientifically assessed. The influence of extreme energy values ​​is smoothed by using a logarithmic function, and the distance attenuation parameter and damping parameter are used to truly reflect the attenuation law of the wave, thereby more accurately assessing the intensity of external disturbances.

[0058] S103, based on the theory of rock rheological damage, introduces a forgetting mechanism to perform time-varying weighted accumulation of the dynamic mining disturbance response index within the historical time window, and calculates the rock mass rheological damage accumulation factor that reflects the historical fatigue state of the rock mass.

[0059] In this embodiment, since the rock mass has memory properties, past disturbances will cause cumulative damage to the rock mass. Therefore, this embodiment of the invention constructs a rock mass rheological damage accumulation factor. The cumulative damage state of rock mass is characterized by the rock mass rheological damage accumulation factor. The calculation method is as follows:

[0060]

[0061] In the formula, Indicates time The rock mass rheological damage accumulation factor has a value range of [0,1]. Indicates the length of the effective traceability time window, for example, 30 days; Representing historical moments The dynamic mining disturbance response index; Indicates the rock mass self-healing rate coefficient; This represents the rock mass's tolerance capacity constant; Represents variables at historical moments.

[0062] Continuing with the above example, the calculation demonstration will proceed as follows:

[0063] Assuming a trace window There was only one disturbance yesterday. The value was 3.93, with no disturbances at other times.

[0064] Let the rock mass self-healing rate coefficient be set. The rock mass tolerance constant is 0.1. It is 10.

[0065] First, calculate the time decay weight: .

[0066] Calculate the weighted disturbance value: .

[0067] Substitute into the main expression:

[0068] .

[0069] This means that although there was no disturbance today, the rock mass still bears about 30% of the accumulated damage due to yesterday's impact. If it is subjected to impact for several consecutive days, this value will continue to approach 1, indicating that the rock mass is close to failure.

[0070] As can be seen from the above process, by calculating the cumulative factor of rock mass rheological damage, the real physical process of rock mass injury, accumulation and partial recovery can be reflected. This can help identify potential danger zones that are currently calm but have been severely damaged in the past, thus avoiding misjudgment of dangerous areas.

[0071] S104. Construct a similarity matching model that includes a damage correction dimension. Calculate the comprehensive matching score based on the differences in static geological characteristics between the area to be supported and historical cases, as well as the differences in the cumulative factor of rock mass rheological damage. Recommend target support parameters based on the comprehensive matching score.

[0072] In this embodiment, the rock mass rheological damage accumulation factor obtained in step S103 is added as a key dimension to the database retrieval.

[0073] The specific calculation expression is as follows:

[0074]

[0075] In the formula, This indicates the current area awaiting support and the first case in the historical case database. The overall matching score of each case This represents the total number of static geological characteristic indicators. It is the first The weight of each static geological characteristic index, Indicates the current area to be supported. A static geological characteristic index, Indicates the first The first historical case Examples of static geological characteristic indicators include uniaxial compressive strength of rock and rock quality indicators; This represents the cumulative rheological damage factor of the rock mass in the area to be supported at the current moment. Indicates the first The cumulative factor of rock mass rheological damage during support implementation in a historical case. It is the penalty weight coefficient for the damage dimension; It is a numerical stability constant, exemplarily 1.

[0076] Calculation example:

[0077] Suppose there are two historical cases, A and B.

[0078] Area to be supported: Static characteristic values The current damage is 0.5. A value of 0.8 indicates a high damage state.

[0079] Case A: Static parameters are highly similar but the conditions are in a low-damage state. The value is 0.5, representing historical damage. Case B: Static parameters are slightly worse but damage is high, (0.1) The value is 0.6, indicating historical damage. The weight is 0.8. The penalty coefficient is 1. It is 10. The value is 1.

[0080] Calculate the matching denominator for case A:

[0081] Static difference squared: Damage difference items: , = ,Score .

[0082] Calculate the matching denominator for case B:

[0083] Static difference squared: Damage difference items: , = ,Score .

[0084] The results showed that although the static geological features of Case A were completely consistent with the area to be supported, the damage factors of Case A and the area to be supported were significantly different due to Case A being in a low damage accumulation state while the area to be supported was in a high damage accumulation state. Therefore, Case A's matching score was much lower than Case B's. Although Case B had slight differences in static geological features compared to the area to be supported, its damage accumulation state was highly consistent with the area to be supported, resulting in a higher overall matching score. Based on this, the system recommended support parameters for Case B. Since Case B implemented greater support resistance under high damage conditions, it effectively ensured the safety of the area to be supported. Finally, the system selected several cases with the highest matching scores, performed a weighted average of their support parameters, and output the recommended support parameters.

[0085] As can be seen from the above process, by using similarity matching that includes damage correction, a precise recommendation combining dynamic and static elements is achieved. This forces the system to pay attention to the hidden damage state of the rock mass, preventing insufficient support resistance due to relying solely on surface geological parameters, and improving the safety of the scheme.

[0086] To further illustrate the effects of this invention, please refer to... Figure 2 and Figure 3 Please provide an explanation.

[0087] Reference Figure 2 The bar chart representing the dynamic mining disturbance response index peaks with the occurrence of disturbance events, while the line graph representing the rock mass rheological damage accumulation factor gradually increases with the continued occurrence of disturbance events. In the early stage when the disturbance is relatively weak, the damage factor rises slowly; as the frequency and intensity of the disturbance increase, the damage factor rises rapidly and continues to accumulate, approaching a critical value. Thereafter, even if the disturbance intensity fluctuates, the damage factor remains at a high level, reflecting the memory effect of irreversible accumulation of rock mass damage. This verifies that the model of this invention can effectively capture the historical cumulative damage state of the rock mass under repeated disturbances.

[0088] Reference Figure 3 Existing technologies cannot detect dynamic damage in the high-disturbance region, resulting in recommended values ​​that are far lower than actual requirements, creating a safety gap. In contrast, this invention closely follows the actual demand curve and automatically increases the recommended support resistance in the high-disturbance region. In the high-disturbance region, the deviation between the support resistance recommended by this invention and the actual demand is significantly smaller than that of existing technologies, effectively eliminating the safety gap.

[0089] It should be noted that those skilled in the art can make various modifications and improvements without departing from the inventive concept, and these all fall within the scope of protection of this invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for intelligent recommendation of mine support schemes using environmental analogy databases, characterized in that: include: Acquire multi-source time-series data of the area to be supported and perform benchmarking preprocessing. The multi-source time-series data includes microseismic monitoring data, static geological data, and historical case database data. Based on the energy values ​​and focal distances of microseismic events in the microseismic monitoring data, and combined with the attenuation characteristics of the geological medium, a dynamic mining disturbance response index reflecting the instantaneous impact intensity of the current mining activity on the area to be supported is calculated, satisfying the following: Indicates time The dynamic mining disturbance response index. This indicates the total number of microseismic events detected within the current time window. Indicates the first The energy value released by this event. Indicates the first The Euclidean distance between the epicenter coordinates of the event and the center point of the area to be supported. Indicates the damping coefficient of the geological medium. Represents the power exponent of geometric decay of energy. Represents the energy normalization coefficient; Based on the theory of rock rheological damage, a forgetting mechanism is introduced to perform time-varying weighted accumulation of the dynamic mining disturbance response index within the historical time window, and the rock mass rheological damage accumulation factor reflecting the historical fatigue state of the rock mass is calculated, satisfying: Indicates time Rock mass rheological damage accumulation factor Indicates the length of the effective traceability time window. Indicates the rock mass self-healing rate coefficient. This represents the rock mass's tolerance capacity constant; A similarity matching model including a damage correction dimension is constructed. Based on the differences in static geological characteristics between the area to be supported and historical cases in the historical case database, as well as the differences in the cumulative factor of rock mass rheological damage, a comprehensive matching score is calculated. Target support parameters are then selected and recommended from the historical case database based on the comprehensive matching score.

2. The intelligent recommendation method for mine support schemes based on environmental analogy databases according to claim 1, characterized in that, The microseismic monitoring data includes the time of occurrence of the microseismic event, the coordinates of the epicenter, and the energy released; the static geological data includes rock quality indicators, uniaxial compressive strength of the rock, and joint orientation information; the historical case database includes geological environment descriptions of historical cases, implemented support parameters, and maintenance records.

3. The intelligent recommendation method for mine support schemes based on environmental analogy databases according to claim 2, characterized in that, The damping coefficient of the geological medium is derived from the rock integrity coefficient, and its value is inversely proportional to the rock quality index; the energy geometric attenuation power exponent is set according to the wave velocity attenuation law of the geological medium.

4. The intelligent recommendation method for mine support schemes based on environmental analogy databases according to claim 1, characterized in that, The self-healing rate coefficient of the rock mass is determined according to the rock type. A smaller value is taken for brittle rocks to characterize strong memory, and a larger value is taken for soft rocks to characterize weak memory. The rock mass tolerance capacity constant is determined by laboratory rock fatigue test.

5. The intelligent recommendation method for mine support schemes based on environmental analogy databases according to claim 1, characterized in that, The calculation of the overall matching score includes: In the formula, This indicates the current area awaiting support and the first case in the historical case database. The overall matching score of each case This represents the total number of static geological characteristic indicators. Indicates the first The weight of each static geological characteristic index, Indicates the current area to be supported. A static geological characteristic index, Indicates the first The first historical case A static geological characteristic index, This represents the cumulative rheological damage factor of the rock mass in the area to be supported at the current moment. Indicates the first The cumulative factor of rock mass rheological damage during support implementation in a historical case. This represents the penalty weight coefficient for the damage dimension. This represents the numerical stability constant.

6. The intelligent recommendation method for mine support schemes based on environmental analogy databases according to claim 5, characterized in that, The weights of the static geological feature indicators are determined by principal component analysis; the penalty weight coefficient of the damage dimension is set to a constant greater than 1.

7. The intelligent recommendation method for mine support schemes based on environmental analogy databases according to claim 1, characterized in that, The benchmarking preprocessing includes: The energy data in the microseismic monitoring data are logarithmically processed to eliminate differences in magnitude; The unstructured geological descriptions in the static geological data are mapped to numerical values ​​using an expert scoring method; A unified time sampling window is set, and the continuously collected data is discretized and aligned according to the time sampling window.

8. The intelligent recommendation method for mine support schemes based on environmental analogy databases according to claim 1, characterized in that, The recommended target support parameters include: selecting a preset number of historical cases with the highest comprehensive matching score, and calculating the weighted average of the support parameters of the historical cases to obtain the final recommended support parameters.

Images