A method, system and device for rapid online detection of ash content in slurry

Abstract

The present invention discloses a method, system and device for rapid online detection of ash content in slurry, and relates to the field of intelligent coal preparation technology. The method comprises: obtaining slurry characteristic data at the current moment; the slurry characteristic data comprises: slurry resistivity and slurry density; inputting the slurry characteristic data at the current moment into the slurry ash prediction model of the current period to obtain the predicted ash value at the current moment; the slurry ash prediction model of the current period is determined by a dynamic update method based on a coal quality characteristic database, slurry characteristic data at historical moments and real ash values ​​at historical moments; the coal quality characteristic database comprises several groups of slurry characteristic data collected experimentally and real ash values ​​collected experimentally. The present invention can improve the detection accuracy and shorten the detection cycle, thereby timely and effectively feeding back information to the flotation system, and then timely adjusting the flotation reagent dosage, and guiding the coal preparation production in real time.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent coal preparation, and in particular to a method, system and device for rapid online detection of ash content in slurry. Background Art

[0002] Under the background of dual carbon, it is a general trend to promote the construction of intelligent coal preparation plants. Intelligent flotation is the last ceiling of intelligent coal preparation. As the most widely used and effective coal preparation process for the separation and recovery of coal slimes below 0.5 mm, the realization of intelligent flotation is of great significance to the clean and efficient use of coal. However, since there is currently no effective, reliable, and real-time rapid detection technology and equipment for ash content in slurry, it has become a technical bottleneck restricting the intelligentization of flotation. In the process of promoting intelligent flotation, the most important and most critical technical link in the current technical field is the online real-time detection technology of ash content in slurry.

[0003] At present, coal preparation plants generally still use the combustion method to manually measure the ash content of ore pulp, which has a long detection cycle, low efficiency, poor sampling representativeness, human errors and other problems. At present, there are three main detection technologies that can be used for ore pulp ash: flotation foam ash measurement technology based on image recognition, X-ray fluorescence detection ore pulp ash meter technology for flotation concentrate, and color difference visual recognition prediction ore pulp ash meter for flotation tailings. However, there are problems with the above three detection methods. For example, poor repeatability, susceptibility to changes in coal quality, high environmental requirements, high cost, complex structure, large system maintenance, long measurement cycle, and difficult to ensure accuracy. Therefore, the above three measurement technologies cannot provide timely and effective feedback information to the flotation system, so that the amount of flotation reagents cannot be adjusted in time to guide coal preparation production in real time. In addition, some detection technologies require the use of radioactive sources, which increases safety hazards. Summary of the invention

[0004] The object of the present invention is to provide a method, system and device for rapid online detection of ash content in slurry, so as to improve detection accuracy and shorten detection cycle.

[0005] To achieve the above object, the present invention provides the following solutions:

[0006] A method for rapid online detection of ash content in slurry, the method comprising:

[0007] Acquire the slurry characteristic data at the current moment; the slurry characteristic data includes: slurry resistivity and slurry density;

[0008] The slurry characteristic data at the current moment is input into the slurry ash prediction model of the current time period to obtain the predicted ash value at the current moment; the slurry ash prediction model of the current time period is determined by a dynamic update method based on a coal quality characteristic database, slurry characteristic data at historical moments and real ash values ​​at historical moments; the coal quality characteristic database includes several groups of experimentally collected slurry characteristic data and experimentally collected real ash values.

[0009] Optionally, the method for determining the slurry ash prediction model for the current period includes:

[0010] Obtaining the slurry characteristic data and the real ash value of the previous period; the previous period and the current period are continuous periods, the current period is the period of the current moment; the previous period includes several historical moments;

[0011] Determine N groups of linear function parameter values ​​according to the slurry characteristic data of the previous period and the real ash value of the previous period; each group of linear function parameter values ​​includes a slope value and an intercept value;

[0012] Using a loop iteration method, a set of optimal function parameter values ​​is determined based on the N sets of primary function parameter values;

[0013] The slurry ash prediction model of the previous time period is updated according to the optimal function parameter value to obtain the slurry ash prediction model of the current time period; when the current time period is a set time period, the slurry ash prediction model of the current time period is determined according to the coal quality characteristic database.

[0014] Optionally, the adopting a loop iteration method to determine a set of optimal function parameter values ​​based on N sets of primary function parameter values ​​specifically includes:

[0015] Taking the intercept value as the horizontal coordinate and the slope value as the vertical coordinate, taking each group of the linear function parameter values ​​as a discrete point, to obtain a candidate point set;

[0016] A discrete point is randomly selected from the candidate point set as the target point, and the remaining discrete points in the candidate point set are used as the preliminary point set corresponding to the target point, so that Nk target points and corresponding preliminary point sets are obtained in total; wherein Nk is the total number of discrete points in the candidate point set, and k is an integer starting from 0;

[0017] Determine the center of the smallest circle containing all discrete points in the candidate point set as the initial optimal point;

[0018] Determine the center of the smallest circle containing all discrete points in any one of the prepared point sets as the second-generation optimal point corresponding to the prepared point set, and obtain a total of Nk second-generation optimal points;

[0019] Calculate the distance between each of the second-generation optimal points and the first-generation optimal point, and determine the target point corresponding to the second-generation optimal point with the largest distance value as the most discrete point;

[0020] The slurry ash content prediction model of the previous period is updated with the linear function parameter value corresponding to the most discrete point to obtain a target prediction model;

[0021] Inputting the slurry characteristic data of the previous period into the target prediction model to obtain the predicted ash value of the previous period;

[0022] Determine whether the sum of the absolute values ​​of the errors between the predicted ash value and the actual ash value at each historical moment in the previous period is less than the set error threshold, and obtain a determination result;

[0023] If the judgment result is no, the most discrete point is removed from the candidate point set, the value of k is updated, and the process returns to the step of "selecting any discrete point from the candidate point set as the target point, and using the remaining discrete points in the candidate point set as the preliminary point set corresponding to the target point, to obtain a total of Nk target points and corresponding preliminary point sets";

[0024] If the judgment result is yes, the linear function parameter value corresponding to the initial optimal point is determined as the optimal function parameter value.

[0025] Optionally, determining N groups of linear function parameter values ​​according to the slurry characteristic data of the previous period and the real ash value of the previous period specifically includes:

[0026] According to the slurry characteristic data at each historical moment in the previous period, respectively calculating the slurry characteristic value at each historical moment in the previous period;

[0027] The pulp characteristic values ​​at any two historical moments in the previous period are respectively used as independent variables, and the two real ash values ​​corresponding to the two pulp characteristic values ​​are respectively used as dependent variables, and a linear function fitting is performed to obtain a total of N groups of linear function parameter values; wherein, n is the total number of historical moments in the previous period.

[0028] Optionally, the true ash value is a pulp ash value determined by a combustion method.

[0029] A rapid online detection system for ash content in slurry, the system comprising:

[0030] The data acquisition module is used to acquire the slurry characteristic data at the current moment; the slurry characteristic data includes: slurry resistivity and slurry density;

[0031] The ash content prediction module is used to input the slurry characteristic data at the current moment into the slurry ash content prediction model of the current time period to obtain the predicted ash content value at the current moment; the slurry ash content prediction model of the current time period is determined by a dynamic update method based on a coal quality characteristic database, slurry characteristic data at historical moments and real ash content values ​​at historical moments; the coal quality characteristic database includes several groups of slurry characteristic data collected experimentally and real ash content values ​​collected experimentally.

[0032] A rapid online detection device for ash content in slurry, the device comprising: a characteristic data collector, a sampling and ash content detector and the above system; the characteristic data collector and the sampling and ash content detector are respectively connected to the system;

[0033] The characteristic data collector is used to collect the slurry characteristic data at the current moment and send the slurry characteristic data to the system;

[0034] The sampling and ash detector is used to sample the ore pulp at a historical moment and detect the real ash value of the sample.

[0035] Optionally, the minimum time interval for the characteristic data collector to collect slurry characteristic data is 10 milliseconds.

[0036] Optionally, the minimum time interval for the sampling and ash detector to perform slurry sampling and ash detection is 1 second.

[0037] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0038] The present invention adopts a dynamic update method to determine the slurry ash prediction model of the current time period based on the coal quality characteristic database, the slurry characteristic data at historical moments and the real ash value at historical moments. By inputting the slurry characteristic data at the current moment into the slurry ash prediction model of the current time period, the predicted ash value at the current moment can be quickly obtained. Compared with the existing slurry ash detection method, the present invention dynamically updates the slurry ash prediction model so that the prediction model can adapt to changes in coal quality, improves the prediction accuracy, and the prediction model can predict the slurry ash value in real time according to the input slurry characteristic data, shortening the prediction cycle. In addition, during the detection process, the present invention does not need to use any radiation source and optical device, so it also has the advantages of safety, reliability and low cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0040] Figure 1 A flow chart of the method for rapid online detection of ash content in slurry provided by the present invention;

[0041] Figure 2 A module diagram of a rapid online detection system for ash content in slurry provided by the present invention;

[0042] Figure 3 A specific flow chart of the method for rapid online detection of ash content in slurry provided by the present invention;

[0043] Figure 4 A schematic diagram of the functional relationship offset provided by the present invention;

[0044] Figure 5 Schematic diagrams of the first to third sample sets provided by the present invention;

[0045] Figure 6 A schematic diagram of the dimensionality reduction method provided by the present invention.

[0046] Explanation of symbols:

[0047] Data acquisition module—1, ash content prediction module—2. DETAILED DESCRIPTION

[0048] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] The object of the present invention is to provide a method, system and device for rapid online detection of ash content in slurry, so as to improve detection accuracy and shorten detection cycle.

[0050] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0051] like Figure 1 As shown, the present invention provides a method for rapid online detection of ash content in slurry, the method comprising:

[0052] Step S1: Acquire the slurry characteristic data at the current moment; the slurry characteristic data includes: slurry resistivity and slurry density.

[0053] Step S2: input the slurry characteristic data at the current moment into the slurry ash prediction model for the current time period to obtain the predicted ash value at the current moment; the slurry ash prediction model for the current time period is determined by a dynamic update method based on a coal quality characteristic database, slurry characteristic data at historical moments and real ash values ​​at historical moments; the coal quality characteristic database, also called an expert database, is a coal quality characteristic database obtained through a large number of preliminary laboratory tests, and is composed of a large number of professional slurry coal quality characteristic data and corresponding ash data, that is, the coal quality characteristic database includes several groups of experimentally collected slurry characteristic data and experimentally collected real ash values.

[0054] Preferably, the true ash value is a pulp ash value determined by a combustion method.

[0055] like Figure 2 As shown, in order to execute the above method to achieve the corresponding functions and technical effects, the present invention also provides a rapid online detection system for ash content in slurry, the system comprising:

[0056] The data acquisition module 1 is used to acquire the slurry characteristic data at the current moment; the slurry characteristic data includes: slurry resistivity and slurry density.

[0057] The ash content prediction module 2 is used to input the slurry characteristic data at the current moment into the slurry ash content prediction model of the current time period to obtain the predicted ash content value at the current moment; the slurry ash content prediction model of the current time period is determined by a dynamic update method based on the coal quality characteristic database, the slurry characteristic data at historical moments and the real ash content values ​​at historical moments; the coal quality characteristic database includes several groups of slurry characteristic data collected experimentally and the real ash content values ​​collected experimentally.

[0058] In addition, the present invention also provides a rapid online detection device for ash content in slurry, the device comprising: a characteristic data collector, a sampling and ash content detector and the above system; the characteristic data collector and the sampling and ash content detector are respectively connected to the system. The characteristic data collector is used to collect the characteristic data of the slurry at the current moment and send the slurry characteristic data to the system; the sampling and ash content detector is used to sample the slurry at the historical moment and detect the real ash content value of the sample.

[0059] Preferably, the minimum time interval for the characteristic data collector to collect slurry characteristic data is 10 milliseconds. The present invention adopts a high-precision coal quality characteristic data collector, which is installed below the slurry liquid surface, and can collect and output characteristic data such as slurry resistivity and slurry density in real time. The collection frequency can reach 100 times per second, and the accuracy can reach 0.1%. It is flexible to install, stable in operation, and easy to maintain. It can stably and reliably perform long-term data collection work in the harsh environment of a coal preparation plant.

[0060] Preferably, the minimum time interval between the sampling and ash detector for slurry sampling and ash detection is 1 second.

[0061] As a specific implementation method, see Figure 3 The present invention establishes a coal quality characteristic database through laboratory experiments. At the beginning, a high-precision coal quality characteristic data collector is used to obtain slurry characteristic data such as slurry concentration, temperature, and real-time flow in real time. The real-time collected slurry characteristic data is processed through a slurry ash prediction model established based on the coal quality characteristic database to obtain the real-time slurry ash content. The collector data collection minimum time interval is 10 milliseconds, which can achieve real-time ash monitoring with a minimum interval of 1 second, thereby realizing real-time online and extremely fast measurement of ash content.

[0062] Specifically, the working principle and working process of the present invention are as follows:

[0063] 1) Obtain slurry characteristic data such as slurry concentration, temperature, and real-time flow rate in real time through a high-precision coal quality characteristic data collector.

[0064] 2) The real-time collected slurry characteristic data is processed through the slurry ash prediction model established based on the coal quality characteristic expert database to obtain the real-time slurry ash content. The minimum time interval for data collection of the collector is 10 milliseconds, and real-time ash monitoring with a minimum interval of 1 second can be achieved, thereby realizing real-time online and extremely fast measurement of ash content.

[0065] The slurry ash content prediction model has a big data learning function, and can perform big data analysis and autonomous learning by accumulating real combustion method ash content data and corresponding slurry characteristic data. According to the error (such as ash content standard deviation) compared with the pre-set error threshold, the slurry ash content prediction model is fine-tuned to improve the prediction accuracy.

[0066] Specifically, since the slurry ash prediction model can be updated dynamically in real time, the slurry ash prediction models in different time periods are different. The method for determining the slurry ash prediction model in the current time period includes:

[0067] Step 1: Obtain the slurry characteristic data of the previous period and the real ash value of the previous period; the previous period and the current period are continuous periods, and the current period is the period where the current moment is located; the previous period includes several historical moments.

[0068] Step 2: Determine N groups of linear function parameter values ​​based on the slurry characteristic data of the previous period and the real ash value of the previous period; each group of linear function parameter values ​​includes a slope value and an intercept value. This step specifically includes:

[0069] Step 2.1: According to the slurry characteristic data at each historical moment in the previous period, respectively calculate the slurry characteristic value at each historical moment in the previous period.

[0070] Step 2.2: Take the slurry characteristic values ​​at any two historical moments in the previous period as independent variables, take the two real ash values ​​corresponding to the two slurry characteristic values ​​as dependent variables, perform linear function fitting, and obtain N sets of linear function parameter values; wherein, n is the total number of historical moments in the previous period.

[0071] Step 3: Using a loop iteration method, determine a set of optimal function parameter values ​​based on the N sets of primary function parameter values. This step specifically includes:

[0072] Step 3.1: Taking the intercept value as the horizontal coordinate and the slope value as the vertical coordinate, taking each group of linear function parameter values ​​as a discrete point, and obtaining a candidate point set.

[0073] Step 3.2: Randomly select a discrete point from the candidate point set as the target point, and use the remaining discrete points in the candidate point set as the preliminary point set corresponding to the target point, to obtain a total of Nk target points and corresponding preliminary point sets; wherein Nk is the total number of discrete points in the candidate point set, and k is an integer starting from 0.

[0074] Step 3.3: Determine the center of the smallest circle that includes all discrete points in the candidate point set as the initial optimal point.

[0075] Step 3.4: Determine the center of the smallest circle that includes all discrete points in any one of the prepared point sets as the second-generation optimal point corresponding to the prepared point set, and obtain a total of Nk second-generation optimal points.

[0076] Step 3.5: Calculate the distance between each of the second-generation optimal points and the first-generation optimal point, and determine the target point corresponding to the second-generation optimal point with the largest distance value as the most discrete point.

[0077] Step 3.6: Update the slurry ash prediction model of the previous period with the linear function parameter value corresponding to the most discrete point to obtain the target prediction model.

[0078] Step 3.7: Input the slurry characteristic data of the previous period into the target prediction model to obtain the predicted ash value of the previous period.

[0079] Step 3.8: Determine whether the sum of the absolute values ​​of the errors between the predicted ash value and the actual ash value at each historical moment in the previous period is less than the set error threshold, and obtain the judgment result.

[0080] Step 3.9: If the judgment result is no, the most discrete point is removed from the candidate point set, the value of k is updated, and the step of "arbitrarily selecting a discrete point from the candidate point set as the target point, and using the remaining discrete points in the candidate point set as the preliminary point set corresponding to the target point, to obtain a total of Nk target points and corresponding preliminary point sets" is returned.

[0081] Step 3.10: If the judgment result is yes, the linear function parameter value corresponding to the initial optimal point is determined as the optimal function parameter value.

[0082] Step 4: Update the slurry ash prediction model of the previous time period according to the optimal function parameter value to obtain the slurry ash prediction model of the current time period; when the current time period is a set time period, determine the slurry ash prediction model of the current time period according to the coal quality characteristic database.

[0083] In practical applications, the above steps can be specifically described as:

[0084] The first step is to establish a database of slurry characteristic data α, γ and the corresponding slurry ash (i.e., the actual ash value) as a coal quality characteristic database through a large number of laboratory experiments.

[0085] In the second step, by analyzing the data in the coal quality characteristic database, it is obtained that there is a corresponding functional relationship between the slurry characteristic data α, γ and the slurry ash value (1), which is used as the ash algorithm model.

[0086] f(α,γ)=Y (1)

[0087] In formula (1), α is the resistivity of the slurry; γ is the density of the slurry; and Y is the ash content of the slurry.

[0088] The third step is to use the slurry characteristic data α and γ collected in real time to give the predicted slurry ash value through the above ash algorithm model, that is, the predicted ash value at the current moment.

[0089] The fourth step is to give the above ash algorithm model a big data learning function so that it can realize the automatic update and iteration of function (1). The ash algorithm model with the big data learning function is called an intelligent ash algorithm model. Using the intelligent ash algorithm model, function (1) can be automatically corrected according to the change of coal quality, and the prediction accuracy of the ash algorithm model can be continuously improved.

[0090] In fact, if Figure 4 As shown, the properties of coal slime feed will fluctuate and change as production progresses, and inevitable errors will occur when collecting data. This results in the correspondence between the actual characteristic data and ash content not being completely consistent with the functional relationship preset based on the coal quality characteristic database, and the actual situation has shifted.

[0091] This is also a common problem of some ash meters such as foam images and tailings color difference on the market, that is, they cannot dynamically learn the changes in raw coal and dynamically update the algorithm. The present invention specially establishes a set of intelligent learning algorithms to solve the problem of raw coal property changes during coal slime flotation. It can perform big data learning through the slurry feature data collected at historical moments and the corresponding real ash value, so as to dynamically learn the changes in raw coal and dynamically update the algorithm.

[0092] From the experimental results, it can be seen that the functional relationship between the characteristic data and the ash content can be approximately regarded as linear, which can be expressed by formula (2):

[0093] AX+B=Y (2)

[0094] Among them, X is the characteristic value, which is used to characterize the slurry characteristic data α and γ, and can be specifically calculated from the slurry characteristic data α and γ; Y is the ash value of the slurry; A is the slope in the linear function parameter, and B is the intercept in the linear function parameter.

[0095] Specifically, the calculation formula of the eigenvalue X is:

[0096]

[0097] Among them, k1 and k2 are both constants.

[0098] With the change of coal quality and the inevitable errors in production and data collection, the predicted ash content function relationship of the coal sample shifts and fluctuates from the original position, so an algorithm is needed to determine its accurate function parameters A and B.

[0099] The characteristic data collected by the characteristic data collector is used as the first sample set, and the combustion method ash value of the corresponding sample measured by the sampling and ash detector installed therein is used as the second sample set. The first and second sample sets are used as input and output respectively, and the values ​​of the function parameters A and B are inversely deduced, and the inferred function parameters A and B are used as the third sample set. Figure 5 .

[0100] Through Figure 6 The dimensionality reduction method shown uses A as the new y-value of the ordinate and B as the new x-value of the abscissa, transforming the original problem of finding the optimal function relationship into the problem of finding the optimal point in a group of discrete points.

[0101] Specifically, the optimal point can be obtained by formula (4):

[0102] MinFunction r,For(B,A)((B-B0) 2 +(A-A0) 2 ≤r 2 ) (4)

[0103] Among them, r is the radius of the minimum circle that contains all discrete points, and (B0, A0) is the center of the minimum circle, which is also the optimal point obtained.

[0104] However, due to the influence of production conditions and errors, some over-discrete and invalid A and B values ​​will appear. Therefore, the optimal function relationship obtained by directly finding the optimal point is not necessarily the optimal function relationship. Therefore, it is necessary to use a cyclic iteration method to eliminate over-discrete and invalid A and B values, and determine the optimal point based on the remaining A and B values, and then determine the optimal function parameters. The specific steps are:

[0105] Step 1: There are N groups of A and B values. Based on all N groups of A and B values, determine an optimal point as the base point (i.e., the first-generation optimal point). Then, based on any N-1 groups of A and B values, determine the second-generation optimal point. A total of N different second-generation optimal points are obtained. The distance between each second-generation optimal point and the base point is determined and sorted. The unselected discrete point corresponding to the second-generation optimal point with the largest distance is taken as the most discrete A and B values.

[0106] Step 2: Take the A and B values ​​represented by this point as function parameters, and use the first sample set as input to get a new set of predicted ash values, which are used as the fourth sample set. The errors of the fourth sample set and the second sample set are compared. If the sum of the absolute values ​​of the errors is greater than the set error threshold, it is considered that the point does not meet the accuracy requirements, and the point is removed from the N groups of A and B. Repeat the first and second steps for the remaining N-1 groups of A and B values.

[0107] Step 3: Through loop iteration, the accuracy of the most discrete point in the remaining points is verified each time. As the points converge more and more, until the most discrete point in a certain step meets the accuracy requirements, then all the remaining points meet the accuracy requirements. At this time, the optimal point is obtained based on all the points that meet the accuracy requirements, and the A and B values ​​corresponding to the optimal point are used as the parameters of the optimal function relationship output to update the slurry ash prediction model.

[0108] The present invention uses the machine learning function of AI, and can adjust and upgrade the ash intelligent algorithm through big data analysis and accumulation, so as to adapt to the changes in the quality of the flotation feed coal, and ensure that the measurement accuracy meets the production needs. Specifically, the present invention classifies and processes the collected data, and then inserts functions and constraints to calculate the parameters of the ash intelligent algorithm model; then substitutes the calculated parameters into the model, uses the coal quality characteristic data for simulation prediction, and compares it with the corresponding combustion method ash data, and eliminates the model parameters with unqualified accuracy through precision control; finally, the model parameters that meet the accuracy requirements are processed to fine-tune the original model parameters, so that the ash intelligent algorithm model is updated as the coal quality changes, and the detection accuracy and stability can be continuously improved.

[0109] In summary, the method, system and device for rapid online detection of ash content in slurry provided by the present invention solve the problems that ash content in slurry cannot be detected in real time, the measurement accuracy is low, the detection equipment cannot learn and control the accuracy by itself, and the ash detection cycle is too long in the existing flotation process. It can simultaneously detect the ash content of feed, clean coal and tailing coal slurry, with an extremely short measurement cycle, high accuracy, machine learning function, adaptability to changes in coal quality, good reproducibility, simple structure, high reliability, low cost, flexible installation method, simple maintenance, no radiation source, no optical device, can work stably for a long time, and maintain good accuracy. Therefore, the present invention can timely and effectively feedback information to the flotation system, timely adjust the amount of flotation reagents, and guide coal preparation production in real time.

[0110] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0111] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A rapid online detection method for ash content in slurry, characterized in that: The method comprises: Acquire the slurry characteristic data at the current moment; the slurry characteristic data includes: slurry resistivity and slurry density; The slurry characteristic data at the current moment is input into the slurry ash prediction model of the current period to obtain the predicted ash value at the current moment; the slurry ash prediction model of the current period is determined by a dynamic update method based on a coal quality characteristic database, slurry characteristic data at historical moments and real ash values ​​at historical moments; the coal quality characteristic database includes several groups of slurry characteristic data collected experimentally and real ash values ​​collected experimentally; The method for determining the slurry ash prediction model for the current period includes: Obtaining the slurry characteristic data and the real ash value of the previous period; the previous period and the current period are continuous periods, the current period is the period of the current moment; the previous period includes several historical moments; Determine N groups of linear function parameter values ​​according to the slurry characteristic data of the previous period and the real ash value of the previous period; each group of linear function parameter values ​​includes a slope value and an intercept value; Using a loop iteration method, a set of optimal function parameter values ​​is determined based on the N sets of primary function parameter values; The slurry ash prediction model of the previous period is updated according to the optimal function parameter value to obtain the slurry ash prediction model of the current period; when the current period is a set period, the slurry ash prediction model of the current period is determined according to the coal quality characteristic database; The method of adopting a loop iteration to determine a set of optimal function parameter values ​​based on N sets of primary function parameter values ​​specifically includes: Taking the intercept value as the horizontal coordinate and the slope value as the vertical coordinate, taking each group of the linear function parameter values ​​as a discrete point, to obtain a candidate point set; A discrete point is randomly selected from the candidate point set as the target point, and the remaining discrete points in the candidate point set are used as the preliminary point set corresponding to the target point, so that Nk target points and corresponding preliminary point sets are obtained in total; wherein Nk is the total number of discrete points in the candidate point set, and k is an integer starting from 0; Determine the center of the smallest circle containing all discrete points in the candidate point set as the initial optimal point; Determine the center of the smallest circle containing all discrete points in any one of the prepared point sets as the second-generation optimal point corresponding to the prepared point set, and obtain a total of Nk second-generation optimal points; Calculate the distance between each of the second-generation optimal points and the first-generation optimal point, and determine the target point corresponding to the second-generation optimal point with the largest distance value as the most discrete point; The slurry ash content prediction model of the previous period is updated with the linear function parameter value corresponding to the most discrete point to obtain a target prediction model; Inputting the slurry characteristic data of the previous period into the target prediction model to obtain the predicted ash value of the previous period; Determine whether the sum of the absolute values ​​of the errors between the predicted ash value and the actual ash value at each historical moment in the previous period is less than the set error threshold, and obtain a determination result; If the judgment result is no, the most discrete point is removed from the candidate point set, the value of k is updated, and the step of "arbitrarily selecting a discrete point from the candidate point set as the target point, and using the remaining discrete points in the candidate point set as the preliminary point set corresponding to the target point, to obtain a total of Nk target points and corresponding preliminary point sets" is returned; If the judgment result is yes, the linear function parameter value corresponding to the initial optimal point is determined as the optimal function parameter value.

2. The method for rapid online detection of ash content in slurry according to claim 1, characterized in that: Determining N groups of linear function parameter values ​​according to the slurry characteristic data of the previous period and the real ash value of the previous period specifically includes: According to the slurry characteristic data at each historical moment in the previous period, respectively calculating the slurry characteristic value at each historical moment in the previous period; The slurry characteristic values ​​at any two historical moments in the previous period are respectively used as independent variables, and the two real ash values ​​corresponding to the two slurry characteristic values ​​are respectively used as dependent variables, and a linear function fitting is performed to obtain a total of N groups of linear function parameter values; wherein n is the total number of historical moments in the previous period.

3. The method for rapid online detection of ash content in slurry according to claim 1, characterized in that: The actual ash value is the ash value of the slurry determined by the combustion method.

4. A rapid online detection system for ash content in slurry, characterized in that: The system comprises: A data acquisition module is used to acquire the slurry characteristic data at the current moment; the slurry characteristic data includes: slurry resistivity and slurry density; An ash content prediction module is used to input the slurry characteristic data at the current moment into the slurry ash content prediction model of the current period to obtain the predicted ash content value at the current moment; the slurry ash content prediction model of the current period is determined by a dynamic update method based on a coal quality characteristic database, slurry characteristic data at historical moments and real ash content values ​​at historical moments; the coal quality characteristic database includes several groups of slurry characteristic data collected experimentally and real ash content values ​​collected experimentally; The method for determining the slurry ash prediction model for the current period includes: Obtaining the slurry characteristic data and the real ash value of the previous period; the previous period and the current period are continuous periods, the current period is the period of the current moment; the previous period includes several historical moments; Determine N groups of linear function parameter values ​​according to the slurry characteristic data of the previous period and the real ash value of the previous period; each group of linear function parameter values ​​includes a slope value and an intercept value; Using a loop iteration method, a set of optimal function parameter values ​​is determined based on the N sets of primary function parameter values; The slurry ash prediction model of the previous period is updated according to the optimal function parameter value to obtain the slurry ash prediction model of the current period; when the current period is a set period, the slurry ash prediction model of the current period is determined according to the coal quality characteristic database; The method of adopting a loop iteration to determine a set of optimal function parameter values ​​based on N sets of primary function parameter values ​​specifically includes: Taking the intercept value as the horizontal coordinate and the slope value as the vertical coordinate, taking each group of the linear function parameter values ​​as a discrete point, to obtain a candidate point set; A discrete point is randomly selected from the candidate point set as the target point, and the remaining discrete points in the candidate point set are used as the preliminary point set corresponding to the target point, so that Nk target points and corresponding preliminary point sets are obtained in total; wherein Nk is the total number of discrete points in the candidate point set, and k is an integer starting from 0; Determine the center of the smallest circle containing all discrete points in the candidate point set as the initial optimal point; Determine the center of the smallest circle containing all discrete points in any one of the prepared point sets as the second-generation optimal point corresponding to the prepared point set, and obtain a total of Nk second-generation optimal points; Calculate the distance between each of the second-generation optimal points and the first-generation optimal point, and determine the target point corresponding to the second-generation optimal point with the largest distance value as the most discrete point; The slurry ash content prediction model of the previous period is updated with the linear function parameter value corresponding to the most discrete point to obtain a target prediction model; Inputting the slurry characteristic data of the previous period into the target prediction model to obtain the predicted ash value of the previous period; Determine whether the sum of the absolute values ​​of the errors between the predicted ash value and the actual ash value at each historical moment in the previous period is less than the set error threshold, and obtain a determination result; If the judgment result is no, the most discrete point is removed from the candidate point set, the value of k is updated, and the step of "arbitrarily selecting a discrete point from the candidate point set as the target point, and using the remaining discrete points in the candidate point set as the preliminary point set corresponding to the target point, to obtain a total of Nk target points and corresponding preliminary point sets" is returned; If the judgment result is yes, the linear function parameter value corresponding to the initial optimal point is determined as the optimal function parameter value.

5. A rapid online detection device for ash content in slurry, characterized in that: The device comprises: a characteristic data collector, a sampling and ash detector and a system as claimed in claim 4; The characteristic data collector and the sampling and ash detector are respectively connected to the system; The characteristic data collector is used to collect the slurry characteristic data at the current moment and send the slurry characteristic data to the system; The sampling and ash detector is used to sample the ore pulp at a historical moment and detect the real ash value of the sample.

6. The rapid online detection device for ash content in slurry according to claim 5, characterized in that: The minimum time interval for the characteristic data collector to collect slurry characteristic data is 10 milliseconds.

7. The rapid online detection device for ash content in slurry according to claim 5, characterized in that: The minimum time interval for the sampling and ash detector to perform slurry sampling and ash detection is 1 second.

Images