Method and system for constructing intelligent water affair digital twinborn model in mining area

Abstract

The invention relates to the technical field of digital twinning, and provides a mining area intelligent water affair digital twinning model construction method and system.The mining area intelligent water affair digital twinning model construction method comprises the steps that the turbidity possible change moment, the turbidity change degree, the suspension possible change moment and the suspension change degree are obtained according to an effluent turbidity sequence, and then the water quality abnormal degree is obtained; obtaining the sludge discharge fault probability according to the continuous high-position condition of the sludge position; the dosing deviation probability is obtained according to the correlation between the water inlet turbid amount sequence and the mud position sequence change; obtaining a water inlet fault probability according to the water inlet flow sequence; according to the matching relation between the water inlet turbidity sequence and the dosing sequence, the dosing deviation probability and the water inlet fault probability are combined to obtain the dosing fault probability; and constructing the mining area intelligent water affair digital twinning model according to the mud discharge fault probability, the water entry fault probability and the dosing fault probability. According to the method, the acquisition frequency of the sensor is adjusted through the fault probability, and the data real-time performance of the water affair digital twin model in the fault state is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of digital twinning, in particular to a mine area intelligent water affair digital twinning model construction method and system. BACKGROUND

[0002] The lack of water resources in the mining area not only affects the sustainable development of the mining area economy, but also exacerbates the contradiction between enterprise production and living water. Under the background of overall tension of water resources in the mining area, mine water as a kind of water resource in the mining area, through the development and efficient use of mine water resources, promoting the optimal allocation and reuse of mine water resources, has important significance for promoting regional economic development.

[0003] Under the background of water resource tension in the mining area, it is necessary to construct a mine area intelligent water affair digital twinning model. The model can realize dynamic monitoring and intelligent control of the whole process of mine water collection, transportation, treatment and reuse through the fusion of Internet of Things perception, edge computing, real-time simulation and intelligent decision-making technology. Not only can it improve the fine and automation level of water resource management, but also can effectively early warn abnormal situations and optimize resource allocation, helping the mining area to realize water saving, green development and safe operation.

[0004] At present, when using the mine area intelligent water affair digital twinning model to intelligently control the mine water treatment process, it is necessary to obtain a plurality of related parameters of the mine water treatment process in real time as data support for intelligent control. In order to reduce the power consumption of the sensor during continuous monitoring, and to avoid data redundancy and edge computing burden caused by full-time high-frequency collection, the sensor often uses a lower collection frequency to meet the daily monitoring needs. However, when the mine water treatment system fails, low-frequency collection cannot meet the needs of real-time state monitoring and real-time data processing, resulting in problems such as difficult fault positioning and data response lag. SUMMARY

[0005] The present application provides a mine area intelligent water affair digital twinning model construction method and system to solve the problem of insufficient data real-time of the digital twinning model under the fault condition of the existing water affair system. The technical scheme adopted is as follows:

[0006] The present application provides a mine area intelligent water affair digital twinning model construction method and system to solve the problem of insufficient data real-time of the digital twinning model under the fault condition of the existing water affair system. The technical scheme adopted is as follows:

[0007] Obtain the water inflow turbidity sequence, water inflow sequence, dosing sequence, mud level sequence, water outflow turbidity sequence and water outflow suspended solids concentration sequence;

[0008] A real-time water quality judgment window is constructed in the outflow turbidity sequence, and according to the difference between the high position condition and the fluctuation condition of the two sides of the elements in the real-time water quality judgment window, the turbidity possible change time, the turbidity change degree, the suspended solids possible change time and the suspended solids change degree are obtained; according to the similar condition of the turbidity possible change time and the suspended solids possible change time, the water quality abnormality degree is obtained in combination with the turbidity change degree and the suspended solids change degree;

[0009] In the sludge level sequence, the sludge discharge fault probability is obtained according to the continuous high position condition of the sludge level in combination with the water quality abnormality degree;

[0010] The inflow turbidity sequence is obtained according to the inflow turbidity sequence and the inflow flow sequence; the dosing deviation probability is obtained according to the correlation between the inflow turbidity sequence and the sludge level sequence; the inflow fault probability is obtained according to the fluctuation condition of the inflow flow sequence in combination with the water quality abnormality degree; the current matching offset degree is obtained according to the matching relationship between the inflow turbidity sequence and the dosing sequence, and the dosing fault probability is obtained in combination with the dosing deviation probability and the inflow fault probability;

[0011] The mine intelligent water affair digital twin model is constructed according to the sludge discharge fault probability, the inflow fault probability and the dosing fault probability.

[0012] Further, the method of constructing the real-time water quality judgment window in the outflow turbidity sequence, and obtaining the turbidity possible change time, the turbidity change degree, the suspended solids possible change time and the suspended solids change degree according to the difference between the high position condition and the fluctuation condition of the two sides of the elements in the real-time water quality judgment window, comprises the following specific method:

[0013] In the outflow turbidity sequence, the last element of the outflow turbidity sequence is taken as the last value of the window, and a window with a length of A is constructed, which is recorded as the real-time water quality judgment window; wherein A is a preset water quality window length;

[0014] The calculation method of the backward water quality abnormality index of the a-th element in the real-time water quality judgment window is as follows:

[0015] B a =C a +(1-C a )×D a

[0016] In the formula, B a is the backward water quality abnormality index of the a-th element in the real-time water quality judgment window; C a is the linear normalization result of the variance of all elements in the interval from the a-th element to the last element in the real-time water quality judgment window; D a is the mean of all elements in the interval from the a-th element to the last element in the real-time water quality judgment window;

[0017] obtaining a forward water quality anomaly index;

[0018] obtaining a difference value between the backward water quality anomaly index and the forward water quality anomaly index of the a-th element in the real-time water quality judgment window, taking a maximum value of the difference value and 0 as a state boundary index of the a-th element in the real-time water quality judgment window;

[0019] taking a time point at which an element with the maximum state boundary index in the real-time water quality judgment window is located as a turbidity possible change time point, and taking the state boundary index of the element at the turbidity possible change time point as a turbidity change degree;

[0020] obtaining the suspended solids possible change time point and the suspended solids change degree.

[0021] Further, the water quality anomaly degree is obtained according to the similar situation of the turbidity possible change time point and the suspended solids possible change time point, and the turbidity change degree and the suspended solids change degree, and the specific obtaining method is as follows:

[0022] E = exp (- |H|) x MIN (F, G)

[0023] In the formula, E is the water quality anomaly degree, H is a time distance between the turbidity possible change time point and the suspended solids possible change time point, F is the turbidity change degree, G is the suspended solids change degree, || is an absolute value function, exp () is an exponential function with a natural constant as a base, and MIN () is a minimum value function.

[0024] Further, the sludge discharge fault probability is obtained according to the continuous high position of the sludge level and the water quality anomaly degree in the sludge level sequence, and the specific method comprises the following steps:

[0025] In the sludge level sequence, a window with a length of K is established as a real-time sludge level judgment window with the last value as a window end, wherein K is a preset sludge level window length.

[0026] The calculation method of the current sludge level anomaly index is as follows:

[0027]

[0028] In the formula, L is the current sludge level anomaly index, N is the number of elements in the real-time sludge level judgment window, M b is the value of the b-th element in the real-time sludge level judgment window, M ′ is the average value of all maximum values in the sludge level sequence, P b is the minimum sludge level value in the interval from the value of the b-th element to the value of the last element in the real-time sludge level judgment window, and softmax () is a weight normalization function. is a ceiling function.

[0029] The product of the current sludge level anomaly index and the water quality anomaly degree is recorded as the sludge discharge fault probability.

[0030] Further, the water inflow turbidity sequence is obtained according to the water inflow turbidity sequence and the water inflow sequence, and the specific method comprises the following steps:

[0031] For any time, the product of the water inflow turbidity at the time and the water inflow is recorded as the water inflow turbidity at the time, and the sequence formed by the water inflow turbidity at all times is recorded as the water inflow turbidity sequence.

[0032] Further, the dosing deviation probability is obtained according to the correlation between the water inflow turbidity sequence and the sludge level sequence, and the specific method comprises the following steps:

[0033] In the water inflow turbidity sequence, a window with a length of Q is constructed at the end of the last element, which is recorded as the water inflow turbidity window; wherein Q is a preset dosing window length;

[0034] In the sludge level sequence, for any element, the difference between the element and the previous element is recorded as the deposition rate of the element; the sequence formed by the deposition rates of all elements is recorded as the sludge level deposition rate sequence; in the sludge level deposition rate sequence, a window with a length of Q is constructed at the end of the last element, which is recorded as the deposition rate window;

[0035] The linear normalization result of the DTW distance between the elements in the water inflow turbidity window and the deposition rate window is taken as the current dosing deviation index;

[0036] The product of the current dosing deviation index and the water quality anomaly degree is recorded as the dosing deviation probability.

[0037] Further, the water inflow fault probability is obtained according to the fluctuation of the water inflow sequence and the water quality anomaly degree, and the specific method comprises the following steps:

[0038] In the water inflow sequence, a window with a length of R is constructed at the end of the last element as a flow fluctuation judgment window, and the linear normalization result of the variance of all elements in the flow fluctuation judgment window is recorded as the current water inflow fluctuation degree; wherein R is a preset flow window length; the product of the current water inflow fluctuation degree and the water quality anomaly degree is recorded as the water inflow fault probability.

[0039] Further, the current matching deviation degree is obtained according to the matching relationship between the water inflow turbidity sequence and the dosing sequence, and the dosing fault probability is obtained by combining the dosing deviation probability and the water inflow fault probability, and the specific method comprises the following steps:

[0040] In the water inflow turbidity sequence, for any element, the difference between the element and the previous element is recorded as the turbidity change rate of the element;

[0041] In the dosing sequence, for any element, the difference between the element and its previous element is recorded as the dosing change rate of the element;

[0042] The current matching offset degree is calculated in the following manner:

[0043]

[0044] In the formula, S is the current matching offset degree; W is the total number of time points in the historical operation of the mine wisdom water system; T c is the dosing deviation probability obtained at the cth time point in the historical operation process of the mine wisdom water system; U c is the turbidity change rate of the corresponding element in the inflow turbidity sequence at the cth time point in the historical operation process of the mine wisdom water system; U is the turbidity change rate of the corresponding element in the inflow turbidity sequence at the current time point; V c is the dosing change rate of the corresponding element in the dosing sequence at the cth time point in the historical operation process of the mine wisdom water system; V is the dosing change rate of the corresponding element in the dosing sequence at the current time point; exp() is an exponential function with a natural constant as the base; || is an absolute value function; softmax() is a weight normalization function;

[0045] According to the current matching offset degree, the dosing deviation probability and the inflow fault probability are combined to obtain the dosing fault probability.

[0046] Further, the dosing fault probability is obtained according to the current matching offset degree, the dosing deviation probability and the inflow fault probability, and the specific acquisition method is:

[0047]

[0048] In the formula, X is the dosing fault probability, Y is the inflow fault probability; T is the dosing deviation probability; S is the current matching offset degree; norm() is a linear normalization function.

[0049] The application also proposes a mine wisdom water digital twin model construction system, which comprises a memory, a processor and a computer program stored in the memory and running on the processor, and the processor implements the steps of the above method when executing the computer program.

[0050] The beneficial effects of the present application are: since the collection frequency of the sensor in the daily monitoring process of the mine wisdom water digital twin model cannot meet the real-time monitoring demand when the wisdom water is abnormal, the collection frequency of the sensor in the area where the fault may occur needs to be upgraded, when the area where the fault may occur is located, first, the quality of the effluent is judged, since when the wisdom water is abnormal, the effluent turbidity and the suspended solids concentration can reflect the treatment effect of the mine water, the present application obtains the water quality abnormality degree by the synchronous maintenance of the high position operation or the irregular fluctuation characteristics of the effluent turbidity and the suspended solids concentration, judges the probability of the current wisdom water abnormality; the operation fault of the mine wisdom water system is mainly caused by the high mud level or the abnormality of the dosing system, the present application first obtains the current mud level abnormality index by the continuous high position of the mud level sequence, and then obtains the mud discharge fault probability by combining the water quality abnormality degree, and judges the mud discharge fault; since the root cause of the dosing deviation mainly comes from the incontrollable water inlet system or the abnormality of the dosing system itself, the present application obtains the water inlet fault probability by the fluctuation of the water inlet flow sequence in combination with the water quality abnormality degree, and obtains the dosing fault probability by the current matching offset degree in combination with the dosing deviation probability and the water inlet fault probability, and distinguishes the faults of the water inlet system and the dosing system. Thus, the present application adjusts the sensor collection frequency in the mine wisdom water digital twin model by the mud discharge fault probability, the water inlet fault probability and the dosing fault probability, and improves the real-time effectiveness of the mine wisdom water digital twin model in monitoring the fault when the wisdom water fails. BRIEF DESCRIPTION OF DRAWINGS

[0051] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description only show some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0052] Figure 1 A mine wisdom water digital twin model construction method flowchart provided by an embodiment of the present application. DETAILED DESCRIPTION

[0053] The technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0054] Please refer to Figure 1It shows a mine area intelligent water digital twin model construction method flow chart provided by one embodiment of the present application, and the method comprises the following steps:

[0055] Step S001, the water turbidity sequence, the water flow sequence, the dosing sequence, the mud level sequence, the water turbidity sequence and the water suspended matter concentration sequence are acquired.

[0056] It should be noted that in the digital twin system, a three-dimensional space model needs to be constructed according to the actual scene, so as to realize the three-dimensional visualization interface, and in the running process of the digital twin system, the data binding, running state mapping and dynamic interaction functions are realized at each sensor position of the three-dimensional space model, so the mine area intelligent water digital twin model is first constructed.

[0057] Specifically, the BIM data and GIS model of the mine water purification treatment system are acquired by the mine engineering department, and at the same time, the device planar layout, pipeline orientation, section size and other information are supplemented by CAD drawings; the structure size and interface point of the dosing device, the sedimentation tank, the valve, the pump and the control cabinet are obtained by using the equipment manual; a plurality of sensors are arranged in the mine water purification treatment system, including but not limited to:

[0058] An online turbidity meter is installed on the upstream of the raw water inlet pipeline; an electromagnetic flowmeter is installed on the water inlet main pipeline; a reagent flowmeter is installed on the reagent dosing pipeline; a radar mud level meter is installed above the sedimentation tank; an online turbidity meter is installed on the water outlet main pipeline; an online suspended matter monitor is arranged in parallel with the water turbidity meter;

[0059] According to the BIM data, the GIS model, the CAD drawing and the equipment manual content, the mine area intelligent water digital twin model is constructed by using Unity, wherein the construction method of the digital twin model is a known technology, and the specific method is not introduced here.

[0060] It should be noted that one of the main purposes of digital twinning is to synchronize the sensor data with the virtual model to realize real-time monitoring and spatial visualization of the physical system, so the sensor data needs to be collected in real time.

[0061] Specifically, the preset collection frequency of all sensors in the intelligent water system is once per minute;

[0062] In the running process of the intelligent water system, the water turbidity sequence is acquired by using the online turbidity meter installed on the upstream of the raw water inlet pipeline;

[0063] The water flow sequence is acquired by using the electromagnetic flowmeter installed on the water inlet main pipeline;

[0064] The dosing sequence is acquired by using the reagent flowmeter installed on the reagent dosing pipeline;

[0065] The mud level sequence is obtained by using a radar mud level meter installed above the sedimentation tank;

[0066] The effluent turbidity sequence is obtained by using an online turbidity meter installed on the effluent main pipe;

[0067] The effluent suspended matter concentration sequence is obtained by using an online suspended matter monitor installed on the effluent main pipe.

[0068] Further, all elements of each obtained sequence are normalized.

[0069] In step S002, a real-time water quality judgment window is constructed in the effluent turbidity sequence, and the turbidity possible change time, the turbidity change degree, the suspended matter possible change time and the suspended matter change degree are obtained according to the difference between the high position and the fluctuation of the elements on both sides of the real-time water quality judgment window; the water quality abnormality degree is obtained according to the similarity of the turbidity possible change time and the suspended matter possible change time, combined with the turbidity change degree and the suspended matter change degree.

[0070] It should be noted that in the operation process of the mine intelligent water affair digital twin model, the collection of the sensor can meet the daily monitoring demand, but when the mine intelligent water affair system fails, the low-frequency collection cannot meet the demand of real-time state monitoring and real-time data processing, resulting in difficult fault positioning and data response lag, so it is necessary to quickly locate the potential fault area through preliminary anomaly detection and system logic judgment, to carry out targeted frequency increase of the related sensor, focus on collecting key data, improve the diagnosis efficiency, and at the same time stabilize the resource consumption and system pressure, to ensure that the digital twin platform can monitor the abnormal state while stably and efficiently operating.

[0071] It should be further noted that in the operation process of the mine intelligent water affair system, the water quality parameter is a direct reflection of the system operation state, which can quickly reflect the treatment effect fluctuation and potential fault risk, so it is necessary to first analyze the change of the water quality parameter to judge whether the fault occurs. Among the effluent parameters of the mine intelligent water affair system, the turbidity and the suspended matter concentration can reflect the treatment effect of the mine water, and when the treatment system operates stably, the effluent turbidity and the suspended matter concentration remain low, and when the mine intelligent water affair system fails, the effluent turbidity and the suspended matter concentration will increase synchronously.

[0072] Specifically, in the effluent turbidity sequence, a real-time water quality judgment window is constructed, and the construction method of the real-time water quality judgment window is as follows: the last element of the effluent turbidity sequence is taken as the last value of the window, a window with a length of A is constructed, and the window is recorded as the real-time water quality judgment window; wherein A is a preset water quality window length, and the embodiment is described by taking A=30 as an example.

[0073] It should be noted that since the mine wisdom water system failure occurs at a certain moment, the turbidity and suspended solids concentration will increase synchronously after the failure occurs, and after increasing to a certain level, with the continuous failure, the system may enter an unstable running state, at this time the turbidity and suspended solids concentration will maintain high running or irregular fluctuation in a higher interval.

[0074] Specifically, the calculation method of the backward water quality abnormality index of the a-th element in the real-time water quality judgment window is:

[0075] B a = C a + (1-C a )xD a

[0076] In the formula, B a is the backward water quality abnormality index of the a-th element in the real-time water quality judgment window; C a is the linear normalization result of the variance of all elements in the interval from the a-th element to the last element in the real-time water quality judgment window, and the normalization object is the variance of all elements from the corresponding last element in all elements of all real-time water quality judgment windows in the historical running process of the mine wisdom water system; D a is the mean of all elements in the interval from the a-th element to the last element in the real-time water quality judgment window.

[0077] It should be noted that the larger C a is, the more likely it is that the turbidity will fluctuate greatly after the a-th element in the real-time water quality judgment window, and the more likely it is to show abnormality; when C a is smaller, but D a is larger, it means that the turbidity may maintain high running in a higher interval after the a-th element in the real-time water quality judgment window, and the more likely it is to show abnormality.

[0078] It should be noted that since the mine wisdom water system failure occurs at a certain moment, the water quality abnormality index before and after the change of water quality will have a large difference, so the forward water quality abnormality index is obtained.

[0079] Specifically, according to the method of obtaining the backward water quality abnormality index, the forward water quality abnormality index of the a-th element in the real-time water quality judgment window is obtained according to all elements in the interval from the a-th element to the first element in the real-time water quality judgment window.

[0080] The difference obtained by subtracting the forward water quality abnormality index from the backward water quality abnormality index of the a-th element in the real-time water quality judgment window is obtained, and the maximum value of the difference and 0 is taken as the state boundary index of the a-th element in the real-time water quality judgment window.

[0081] It needs to be explained that the greater the state boundary index of the a-th element in the real-time water quality judgment window, the more likely it is that the water quality starts to change at the time when the element is located.

[0082] Further, the time when the element with the maximum state boundary index in the real-time water quality judgment window is located is recorded as the turbidity possible change time; and the state boundary index of the element at the turbidity possible change time is recorded as the turbidity change degree.

[0083] It needs to be explained that when the mine area intelligent water system fails to cause the water quality to change, since the effluent turbidity and the suspended matter concentration both directly reflect the change of the particulate matter content in the water, the changes of the effluent turbidity and the suspended matter concentration will maintain synchronicity.

[0084] Further, according to the turbidity possible change time and the turbidity change degree acquisition method, the suspended possible change time and the suspended change degree are acquired according to the effluent suspended matter concentration sequence;

[0085] The calculation method of the water quality abnormality degree is:

[0086] E = exp(-|H|) x MIN(F, G)

[0087] In the formula, E is the water quality abnormality degree; H is the time distance between the turbidity possible change time and the suspended possible change time; F is the turbidity change degree; G is the suspended change degree; || is the absolute value function; exp() is the exponential function with the natural constant as the base; and MIN() is the minimum value function.

[0088] It needs to be explained that the greater exp(-|H|), the closer the turbidity possible change time and the suspended possible change time, the water quality abnormality degree is higher; and the greater MIN(F, G), the greater the change of the turbidity and the suspended matter concentration, and the greater the degree of the water quality abnormality change.

[0089] Step S003, in the sludge level sequence, the sludge discharge fault probability is obtained according to the continuous high sludge level and the water quality abnormality degree.

[0090] It needs to be explained that the operation fault of the mine area intelligent water system is mainly caused by the high sludge level or the abnormality of the dosing system. The high sludge level will cause the sludge-water interface of the sedimentation tank to move upwards, so that the solid-liquid separation effect is poor, a large amount of suspended matter enters the effluent system with the water flow, and the turbidity and the suspended matter concentration exceed the standard; and if the dosing pump of the dosing system fails, the flocculation effect will be poor, the flocculation is insufficient, and the sedimentation and filtration effects are also affected. Since the fault reasons are different, the ranges of the sensors with different sampling frequencies are different, so it is necessary to judge the causes of the fault.

[0091] It needs to be further explained that firstly the possibility of failure caused by the sludge level is analyzed. When the sludge level is maintained within a reasonable range, the sludge discharge system operates normally, the sludge-water interface in the pool is stable, the correlation between the sludge level and the abnormal degree of water quality is weak, and the abnormal degree of water quality is basically maintained stable. When the sludge level is too high, the sedimentation capacity decreases sharply, and the turbidity and suspended solids concentration of effluent increase sharply, which directly leads to the significant deterioration of water quality index.

[0092] Specifically, in the sludge level sequence, the last value is taken as the end of the window, a window with a length of K is established as a real-time sludge level judgment window; wherein K is a preset sludge level window length, and the embodiment is described taking K = 30 as an example.

[0093] It needs to be noted that in the intelligent water system in the mining area, in order to ensure the drug efficacy and system efficiency, the fixed sludge level threshold discharge mode is often used to control the sludge level, and when the sludge level reaches a certain position, the sludge discharge action is triggered, so by comparing the current sludge level with the peak value of the historical sludge level, it is determined whether the current sludge level is continuously too high.

[0094] Specifically, the calculation method of the current sludge level anomaly index is as follows:

[0095]

[0096] In the formula, L is the current sludge level anomaly index; N is the number of elements in the real-time sludge level judgment window; M b is the value of the bth element in the real-time sludge level judgment window; M ′ is the average of all maximum values in the sludge level sequence; P b is the minimum sludge level value in the interval from the bth element value to the last element value in the real-time sludge level judgment window; softmax() is a weight normalization function, and the normalization object is the minimum sludge level value in the interval from each element to the last element value in the real-time sludge level judgment window; is a rounding up function.

[0097] It needs to be noted that when is not 0, it means that the sludge level represented by the bth element in the real-time sludge level judgment window has exceeded the general sludge discharge level, and an abnormality may occur; the larger softmax(M b ), the more the sludge level represented by the bth element in the real-time sludge level judgment window does not decrease significantly by the current time, and the bth element in the real-time sludge level judgment window can better reflect the continuous state of the sludge level at the current time. When the water quality decreases due to the high sludge level, the high sludge level is the cause and the water quality decrease is the result. Therefore, when the water quality is abnormal, the sludge level has been continuously high, so if the sludge level is continuously abnormal, it means that it is more likely to be a sludge discharge failure.

[0098] Further, the product of the current sludge level anomaly index and the water quality anomaly degree is denoted as the sludge discharge fault probability.

[0099] In step S004, the inflow turbidity sequence and the inflow flow sequence are obtained, the dosing deviation probability is obtained according to the correlation between the inflow turbidity sequence and the sludge level sequence, the inflow fault probability is obtained according to the fluctuation of the inflow flow sequence in combination with the water quality anomaly degree, and the current matching deviation degree is obtained according to the matching relationship between the inflow turbidity sequence and the dosing sequence in combination with the dosing deviation probability and the inflow fault probability to obtain the dosing fault probability.

[0100] It should be noted that when the sludge discharge fault probability is low and the water quality anomaly degree is high, it is very likely that the dosing fault is caused. When the reagent is insufficient, excessive, improperly proportioned or unevenly added, the flocculation effect is poor, which further affects the sedimentation efficiency and the effluent water quality.

[0101] It should be further noted that the turbidity reflects the content of suspended particles and colloidal substances in the inflow, and the reagent is mainly added according to the turbidity of the inflow. When the turbidity of the inflow increases, the amount of reagent added needs to be increased to ensure that enough reagent reacts with particles, so that the flocculation particle size increases and the specific gravity increases, which is beneficial to the acceleration of particle settlement in the sedimentation tank. Therefore, the sludge deposition rate usually increases with the inflow turbidity. When the positive correlation is low, it indicates that the dosing system has a fault.

[0102] Specifically, for any time, the product of the inflow turbidity and the inflow flow at that time is denoted as the inflow turbidity at that time, and the sequence formed by the inflow turbidity at all times is denoted as the inflow turbidity sequence. In the inflow turbidity sequence, a window with a length of Q is constructed with the last element as the end, which is denoted as the inflow turbidity window. Wherein, Q is a preset dosing window length, and the embodiment is described by taking Q = 20 as an example.

[0103] In the sludge level sequence, for any element, the difference between the element and its previous element is denoted as the deposition rate of the element. The sequence formed by the deposition rates of all elements is denoted as the sludge level deposition rate sequence. In the sludge level deposition rate sequence, a window with a length of Q is constructed with the last element as the end, which is denoted as the deposition rate window.

[0104] The linear normalized result of the DTW distance between the elements in the inflow turbidity window and the deposition rate window is used as the current dosing deviation index. Wherein, the object of normalization is the DTW distance between the elements in the inflow turbidity window and the deposition rate window obtained at each time during the historical operation of the mine intelligent water management system.

[0105] The product of the current dosing deviation index and the water quality abnormality degree is denoted as a dosing deviation probability.

[0106] It should be noted that in the intelligent control process, the root cause of the dosing deviation mainly comes from the out-of-control of the water inlet system or the abnormality of the dosing system itself. In order to improve the efficiency of subsequent fault monitoring and the accuracy of fault positioning, it is necessary to further trace back and judge the source of the deviation.

[0107] Specifically, in the water inlet flow sequence, a window with a length of R ending with the last element is constructed as a flow fluctuation judgment window, and the linear normalization result of the variance of all elements in the flow fluctuation judgment window is denoted as the current water inlet fluctuation degree; wherein R is a preset flow window length, and the embodiment is described taking R=5 as an example; the product of the current water inlet fluctuation degree and the water quality abnormality degree is denoted as a water inlet fault probability; wherein the linear normalization object is the variance of all elements in each flow fluctuation judgment window constructed in the historical operation process of the mine intelligent water affairs system.

[0108] It should be noted that if the current water inlet fluctuation degree is high, it indicates that the water inlet flow is abnormally fluctuating, which causes the dosing system to fail to match the reagent dosage in time, thereby causing the dosing deviation.

[0109] It should be further noted that if the water inlet fault probability is low but the dosing deviation is high, it is necessary to determine whether the dosing system has failed, causing the change in the dosing amount and the water inlet turbidity to deviate. In the operation process of the mine intelligent water affairs, since the dosing rate is adjusted in real time according to the change in the water inlet turbidity, it is necessary to judge the matching relationship between the water inlet turbidity sequence and the dosing sequence.

[0110] Specifically, in the water inlet turbidity sequence, for any element, the difference between the element and its previous element is denoted as the turbidity change rate of the element; it should be noted that the turbidity change rate of the first element in the water inlet turbidity sequence is 0;

[0111] In the dosing sequence, for any element, the difference between the element and its previous element is denoted as the dosing change rate of the element; it should be noted that the dosing change rate of the first element in the dosing sequence is 0;

[0112] The calculation method of the current matching offset degree is:

[0113]

[0114] In the formula, S is the current matching offset degree; W is the total number of times of the historical operation of the mine intelligent water affairs system; T c is the dosing deviation probability obtained at the cth time in the historical operation process of the mine intelligent water affairs system; U cis the turbidity change rate of the corresponding element in the water inflow turbidity sequence at the cth moment in the historical running process of the mine area intelligent water system; U is the turbidity change rate of the corresponding element in the water inflow turbidity sequence at the current moment; V c is the dosing change rate of the corresponding element in the dosing sequence at the cth moment in the historical running process of the mine area intelligent water system; V is the dosing change rate of the corresponding element in the dosing sequence at the current moment; exp() is an exponential function with a natural constant as the base; || is an absolute value function; softmax() is a weight normalization function, and the normalization object is

[0115] It should be noted that when exp(-T c ) is greater, it means that the probability of dosing deviation at the cth moment in the historical running process of the mine area intelligent water system is smaller, and the reference of the data at this moment is stronger; |U c -U| is smaller, which means that the turbidity change rate at the cth moment in the historical running process of the mine area intelligent water system is more similar to the current moment, and the corresponding dosing amount change is also similar; exp(-|V c -V|) is greater, which means that the change of the dosing amount at the current moment deviates to a greater extent, and the dosing system is more likely to fail.

[0116] It should be further noted that when the water inflow failure probability is low, but the dosing deviation is high, if the current matching offset degree of dosing is higher, it means that the dosing system fails.

[0117] Specifically, the calculation method of the dosing failure probability is:

[0118]

[0119] In the formula, X is the dosing failure probability, Y is the water inflow failure probability; T is the dosing deviation probability; S is the current matching offset degree; norm() is a linear normalization function, and the normalization object is

[0120] Step S005, constructing the mine area intelligent water digital twin model according to the sludge discharge failure probability, the water inflow failure probability and the dosing failure probability.

[0121] It should be noted that after obtaining the sludge discharge failure probability, the water inflow failure probability and the dosing failure probability, the preliminary positioning of the current failure occurrence area has been realized. In order to realize real-time state monitoring and real-time data processing of the failure, and then realize accurate positioning of the failure, it is necessary to improve the sampling frequency of the corresponding area sensor.

[0122] Specifically, the sampling frequency of all sensors in the corresponding area of the sludge discharge system is adjusted by using the sludge discharge fault probability.

[0123] The sampling frequency of all sensors in the corresponding area of the water inlet system is adjusted by using the water inlet fault probability.

[0124] The sampling frequency of all sensors in the corresponding area of the water inlet system is adjusted by using the water inlet fault probability.

[0125] Taking the flow sensor of the water inlet system as an example, the adjusted sampling frequency of the flow sensor of the water inlet system is:

[0126]

[0127] In the formula, Z ′ is the adjusted sampling frequency of the flow sensor of the water inlet system; Y is the water inlet fault probability; and a is a preset adjustment coefficient, which is taken as 0.1 in the embodiment. is a ceiling function.

[0128] It should be noted that the flow sensor of the water inlet system will collect data at a frequency of Z ′ times per minute.

[0129] Further, the corresponding data is collected according to the adjusted sensor collection frequency, and is updated and visually displayed in the digital twin model of the mine area intelligent water affair; it should be noted that the collection frequency is adjusted every minute during the operation of the mine area intelligent water affair, so as to complete the construction of the digital twin model of the mine area intelligent water affair.

[0130] It should be noted that by adjusting the sampling frequency of all sensors in the corresponding range of each system, effective real-time monitoring of faults is realized, which is conducive to subsequent accurate positioning of faults, and at the same time, the repair result is fed back and updated in time through the digital twin model of the mine area intelligent water affair when the fault is repaired.

[0131] The embodiment adopts an exp(-MX) model to present an inverse proportional relationship and normalization processing, MX is an input of the model, and the implementer can set an inverse proportional function and a normalization function according to actual conditions.

[0132] Another embodiment of the present application provides a mine area intelligent water affair digital twin model construction system, which comprises a memory, a processor and a computer program stored in the memory and running on the processor, and the processor implements the method steps S001 to S005 when executing the computer program.

[0133] The above merely provides the preferred embodiment of the present application, and is not used to limit the present application, and any modification, equivalent replacement, improvement, etc. made within the principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for constructing a digital twin model of smart water management in a mining area, characterized in that, The method includes the following steps: Obtain the influent turbidity sequence, influent flow rate sequence, chemical dosing sequence, sludge level sequence, effluent turbidity sequence, and effluent suspended solids concentration sequence; A real-time water quality judgment window is constructed in the effluent turbidity sequence. Based on the differences in the high position and fluctuation of elements on both sides of the real-time water quality judgment window, the possible time of turbidity change, the degree of turbidity change, the possible time of suspension change, and the degree of suspension change are obtained. Based on the similarity of the possible time of turbidity change and the possible time of suspension change, the degree of water quality anomaly is obtained by combining the degree of turbidity change and the degree of suspension change. In the sludge level sequence, the probability of sludge discharge failure is obtained based on the sustained high sludge level and the degree of water quality abnormality. Based on the influent turbidity sequence and the influent flow rate sequence, the influent turbidity sequence is obtained; based on the correlation between the changes in the influent turbidity sequence and the sludge level sequence, the dosing deviation probability is obtained; based on the fluctuation of the influent flow rate sequence and the degree of water quality anomaly, the influent failure probability is obtained; based on the matching relationship between the influent turbidity sequence and the dosing sequence, the current matching deviation degree is obtained, and based on the dosing deviation probability and the influent failure probability, the dosing failure probability is obtained. A digital twin model of smart water management in the mining area was constructed based on the probabilities of sludge discharge failure, water inlet failure, and chemical dosing failure.

2. The method for constructing a digital twin model of smart water management in a mining area according to claim 1, characterized in that, The method for constructing a real-time water quality judgment window in the effluent turbidity sequence, and obtaining the possible time of turbidity change, the degree of turbidity change, the possible time of suspension change, and the degree of suspension change based on the differences in the high position and fluctuation of elements on both sides of the real-time water quality judgment window, includes the following specific methods: In the effluent turbidity sequence, the last element of the effluent turbidity sequence is used as the last value of the window to construct a window of length A, which is denoted as the real-time water quality judgment window; where A is the preset water quality window length. The calculation method for the backward water quality anomaly index of the a-th element in the real-time water quality assessment window is as follows: B a =C a +(1-C a )×D a In the formula, B a C represents the backward water quality anomaly index of the a-th element in the real-time water quality assessment window. a D is the linearly normalized result of the variance of all elements within the interval from the a-th element to the last element in the real-time water quality judgment window; a It is the mean of all elements in the interval from the a-th element to the last element in the real-time water quality judgment window; Obtain the forward water quality anomaly index; Obtain the difference between the backward water quality anomaly index and the forward water quality anomaly index of the a-th element in the real-time water quality judgment window, and take the maximum value between this difference and 0 as the state boundary index of the a-th element in the real-time water quality judgment window. The moment when the element with the largest state boundary index in the real-time water quality judgment window is located is recorded as the moment when turbidity may change; the state boundary index of the element at the moment when turbidity may change is recorded as the degree of turbidity change. Obtain the possible times and degrees of change in the suspension.

3. The method for constructing a digital twin model of smart water management in a mining area according to claim 1, characterized in that, The degree of water quality abnormality is determined by combining the similarities between the possible times when turbidity and suspension may change, and the degree of change in turbidity and suspension. The specific method for obtaining this information is as follows: E = exp(-|H|) × MIN(F,G) In the formula, E represents the degree of water quality abnormality; H represents the time distance between the possible change in turbidity and the possible change in suspended matter; F represents the degree of turbidity change; G represents the degree of suspended matter change; || represents the absolute value function; exp() represents the exponential function with the natural constant as the base; and MIN() represents the minimum value function.

4. The method for constructing a digital twin model of smart water management in a mining area according to claim 1, characterized in that, The method for determining the probability of sludge discharge failure based on the sustained high sludge level in the sludge level sequence, combined with the degree of water quality anomaly, includes the following specific methods: In the mud level sequence, a window of length K is created with the last value as the end of the window, which serves as the real-time mud level judgment window; where K is the preset mud level window length. The current mud level anomaly index is calculated as follows: In the formula, L is the current mud level anomaly index; N is the number of elements in the real-time mud level judgment window; M b M represents the b-th element value in the real-time mud level judgment window; ′ P is the mean of all maxima in the mud level sequence; b The minimum mud level value within the interval from the b-th element value to the last element value in the real-time mud level judgment window; softmax() is the weight normalization function; It is a rounding function; The product of the current sludge level anomaly index and the degree of water quality anomaly is denoted as the sludge discharge failure probability.

5. The method for constructing a digital twin model of smart water management in a mining area according to claim 1, characterized in that, The specific method for obtaining the influent turbidity sequence based on the influent turbidity sequence and the influent flow rate sequence is as follows: For any given moment, the product of the influent turbidity and the influent flow rate at that moment is denoted as the influent turbidity at that moment, and the sequence of influent turbidity at all moments is denoted as the influent turbidity sequence.

6. The method for constructing a digital twin model of smart water management in a mining area according to claim 1, characterized in that, The method for obtaining the dosing deviation probability based on the correlation between changes in influent turbidity sequence and sediment level sequence includes the following specific methods: In the influent turbidity sequence, a window of length Q is constructed with the last element as the end, denoted as the influent turbidity window; where Q is the preset dosing window length. In the mud bed sequence, for any element, the difference between that element and the previous element is denoted as the deposition rate of that element; the sequence of deposition rates of all elements is denoted as the mud bed deposition rate sequence; in the mud bed deposition rate sequence, a window of length Q is constructed with the last element as the end, and is denoted as the deposition rate window. The linearly normalized result of the DTW distance between the elements in the influent turbidity window and the sedimentation rate window is used as the current dosing deviation index. The product of the current dosing deviation index and the degree of water quality abnormality is denoted as the dosing deviation probability.

7. The method for constructing a digital twin model of smart water management in a mining area according to claim 1, characterized in that, The method for determining the probability of water inflow failure based on fluctuations in the inflow sequence and the degree of water quality anomaly includes the following specific methods: In the inflow sequence, a window of length R is constructed with the last element as the end point, which serves as the flow fluctuation judgment window. The linear normalization result of the variance of all elements within the flow fluctuation judgment window is denoted as the current inflow fluctuation degree; where R is the preset flow window length; the product of the current inflow fluctuation degree and the water quality anomaly degree is denoted as the inflow failure probability.

8. The method for constructing a digital twin model of smart water management in a mining area according to claim 1, characterized in that, The method for determining the current matching deviation based on the matching relationship between the influent turbidity sequence and the dosing sequence, and then combining the dosing deviation probability and the influent failure probability to obtain the dosing failure probability, includes the following specific methods: In the turbidity sequence of influent, for any element, the difference between the element and the previous element is denoted as the turbidity change rate of that element. In the dosing sequence, for any element, the difference between that element and the previous element is denoted as the dosing rate of that element. The current matching offset is calculated as follows: In the formula, S represents the current matching offset; W represents the total number of historical operation moments of the mining area's smart water management system; T c U represents the probability of dosing deviation obtained at the c-th moment during the historical operation of the intelligent water management system in the mining area. c The turbidity change rate of the corresponding element in the influent turbidity sequence at the c-th moment during the historical operation of the smart water system in the mining area; U represents the rate of change of turbidity of the corresponding element in the influent turbidity sequence at the current moment; V c Let V be the rate of change of dosing for the corresponding element in the dosing sequence at the c-th moment during the historical operation of the intelligent water management system in the mining area; V is the rate of change of dosing for the corresponding element in the dosing sequence at the current moment; exp() is an exponential function with the natural constant as the base; || is the absolute value function; softmax() is the weight normalization function; Based on the current matching offset, combined with the probability of dosing deviation and the probability of water ingress failure, the probability of dosing failure is obtained.

9. A method for constructing a digital twin model of smart water management in a mining area according to claim 8, characterized in that, The dosing failure probability is obtained by combining the current matching offset degree with the dosing deviation probability and the water inlet failure probability. The specific method for obtaining this probability is as follows: In the formula, X is the probability of dosing failure, Y is the probability of water inlet failure, T is the probability of dosing deviation, S is the current matching offset, and norm() is the linear normalization function.

10. A digital twin model construction system for smart water management in mining areas, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method for constructing a digital twin model of smart water management in a mining area as described in any one of claims 1-9.

Images