An intelligent detection method for pipeline slurry density

Abstract

The present invention relates to an intelligent detection method for pipeline slurry density, which belongs to the technical field of slurry density measurement and solves the problem of low accuracy of existing pipeline slurry density detection methods. The method comprises: real-time collection of the pressure value in the pipeline connected to the slurry pump outlet, the slurry flow value at the slurry pump outlet, the liquid level of the slurry, the frequency and current of the slurry pump motor; inputting the slurry flow value into a trained slurry pump outlet pressure prediction model, and obtaining the slurry pump outlet pressure prediction value after processing; obtaining the pipeline slurry density prediction value based on the slurry pump outlet pressure prediction value and the pipeline pressure value; inputting the slurry flow value, the liquid level of the slurry, the frequency and current of the slurry pump motor into a trained pipeline slurry density compensation model, and obtaining the pipeline slurry density error prediction value after processing; compensating the pipeline slurry density prediction value with the pipeline slurry density error prediction value to obtain the final pipeline slurry density detection result.

Description

Technical Field

[0001] The present invention relates to the technical field of slurry density measurement, and in particular to an intelligent detection method for pipeline slurry density. Background Art

[0002] Flotation is one of the most commonly used mineral processing methods. It is a process of separating minerals from the pulp with the help of the buoyancy of bubbles according to the differences in the physical and chemical properties of the mineral surface. In the flotation process, the main process factors affecting the flotation effect are feed particle size, pulp density, flotation time, reagent system, gas velocity, flotation process and flotation equipment type. Among them, pulp density plays an extremely important role in the entire mineral processing process. It directly affects various technical and economic indicators of the mineral processing process, such as mineral recovery rate, concentrate quality, reagent dosage, flotation machine production capacity, flotation time and energy consumption. In addition, pulp density also has specific requirements in magnetic separation and gravity separation operations, especially for grinding and filtration and dehydration operations. Therefore, it is usually necessary to regularly detect pulp density in industrial production. Only by accurately measuring pulp density can the pulp density be better adjusted and controlled to ensure the stability of the mineral processing production process, thereby improving the concentrate quality and economic benefits. For a long time, reliable and real-time detection of pulp density has been a very difficult problem. The main detection methods currently include density pot detection, drying method detection, nuclear density meter, ultrasonic density meter and differential pressure density meter.

[0003] Density pot and drying method are both manually operated, and the frequency of manual sampling is generally once an hour. The sampling data has a long time interval for real-time process control of production. Although this detection method is widely used, it has high labor intensity and serious lag in data detection, which limits the real-time guidance significance of slurry density for industrial production processes, resulting in unstable product quality and high consumption.

[0004] Both nuclear density meters and ultrasonic density meters can achieve non-contact detection, but nuclear density meters have problems such as nuclear radiation waste disposal. Considering factors such as safety and management, nuclear density meters are difficult to be widely promoted and used in the mineral processing industry; ultrasonic density meters have a high failure rate in industry because they require the sensor to be fully sealed, wear-resistant and highly corrosive. The long-term stability and reliability of this detection method cannot fully meet the mineral processing needs.

[0005] In recent years, differential pressure densitometers have been used more and more widely. They are mainly designed based on the relationship between the static pressure of the slurry in the pipeline, the slurry density and the height difference of the pressure taking point. However, since the dynamic process of the flowing slurry is not considered in its design, and the slurry density error cannot be corrected and compensated, the accuracy of the differential pressure densitometer needs to be further improved to meet the requirements of the mineral processing production process. Therefore, it is extremely necessary to develop an accurate and stable method for online real-time detection of slurry density. Summary of the invention

[0006] In view of the above analysis, an embodiment of the present invention aims to provide a pipeline slurry density intelligent detection method to solve the problem of low accuracy of existing pipeline slurry density detection methods.

[0007] The present invention discloses a pipeline slurry density intelligent detection method, comprising:

[0008] Real-time data collection, including: the pressure value in the pipeline connected to the slurry pump outlet, the slurry flow value at the slurry pump outlet, the slurry liquid level, the frequency and current of the slurry pump motor;

[0009] The real-time collected slurry flow value is input into the trained slurry pump outlet pressure prediction model, and the slurry pump outlet pressure prediction value is obtained after processing;

[0010] Based on the predicted value of the slurry pump outlet pressure and the pressure value in the pipeline, the predicted value of the pipeline slurry density is obtained;

[0011] The real-time collected slurry flow value, slurry liquid level, frequency and current of the slurry pump motor are input into the trained pipeline slurry density compensation model, and the pipeline slurry density error prediction value is obtained after processing;

[0012] The pipeline slurry density prediction value is used to compensate the pipeline slurry density prediction value to obtain a final pipeline slurry density detection result.

[0013] On the basis of the above scheme, the present invention also makes the following improvements:

[0014] Further, the slurry pump outlet pressure prediction model is trained in the following manner:

[0015] Acquire a first sample set, wherein each set of first sample data in the first sample set includes: a slurry flow value at a slurry pump outlet, a pressure value in a pipeline connected to the slurry pump outlet, and a measured value of pipeline slurry density;

[0016] Based on the measured value of pipeline slurry density and the pressure value in the pipeline in each group of first sample data, the standard value of pressure at the outlet of the slurry pump is obtained;

[0017] The slurry flow value at the slurry pump outlet in each group of first sample data is used as input and the pressure standard value at the slurry pump outlet is used as a label, the slurry pump outlet pressure prediction model is trained, the structure and parameters of the slurry pump outlet pressure prediction model are determined, and a trained slurry pump outlet pressure prediction model is obtained.

[0018] Furthermore, in the slurry pump outlet pressure prediction model, the structure and parameters of the slurry pump outlet pressure prediction model are determined based on the LS least squares method.

[0019] Further, the structure and parameters of the slurry pump outlet pressure prediction model are determined in the following manner:

[0020] Predicted pressure value p at the slurry pump outlet ef The slurry flow rate Q at the outlet of the slurry pump satisfies:

[0021] p ef =aQ 2 +b (1)

[0022] The standard pressure value and predicted pressure value at the slurry pump outlet meet the following requirements:

[0023]

[0024] Where n represents the number of groups of the first sample data in the first sample set; p er (i) represents the standard pressure value at the outlet of the slurry pump in the first sample data of the i-th group, p ef (i) represents the predicted pressure value at the outlet of the slurry pump in the first sample data of the i-th group;

[0025] Based on the LS least squares method, the minimum value of J(a,b) is taken as the goal, and the parameters a and b are estimated to obtain the estimated value a of parameter a. * , the estimated value of parameter b * ;

[0026] The trained slurry pump outlet pressure prediction model is:

[0027] p ef =a * Q 2 +b * (3).

[0028] Furthermore, the standard pressure value at the outlet of the slurry pump is obtained according to the following method:

[0029] p er (i) = ρ er (i)gΔH+p c (i) (4)

[0030] Among them, ρer (i) represents the measured value of pipeline slurry density in the first sample data of the i-th group, p c (i) represents the pressure value in the pipeline connected to the slurry pump outlet in the i-th group of first sample data, and ΔH represents the height difference between the slurry pump outlet and the measurement position of the pressure value in the pipeline.

[0031] Further, the pipeline slurry density compensation model is trained in the following manner:

[0032] Acquire a second sample set, wherein each set of second sample data in the second sample set includes: a measured value of pipeline slurry density, a pressure value in a pipeline connected to a slurry pump outlet, a slurry flow value at a slurry pump outlet, a liquid level of slurry, and a frequency and current of a slurry pump motor;

[0033] Inputting the slurry flow rate value at the slurry pump outlet in each group of second sample data into the trained slurry pump outlet pressure prediction model, and obtaining the corresponding slurry pump outlet pressure prediction value after processing;

[0034] Based on the predicted value of the slurry pump outlet pressure and the pressure value in the pipeline corresponding to each set of second sample data, the corresponding predicted value of the pipeline slurry density is obtained;

[0035] The slurry flow value at the slurry pump outlet, the liquid level of the slurry, the frequency and current of the slurry pump motor in each group of second sample data are used as input, and the difference between the pipeline slurry density prediction value and the pipeline slurry density actual value is used as a label to train the pipeline slurry density compensation model to obtain a trained pipeline slurry density compensation model.

[0036] Further, the predicted value of pipeline slurry density ρ is calculated according to the following formula: f ,include:

[0037]

[0038] Among them, p ef represents the predicted pressure at the outlet of the slurry pump, p c It indicates the pressure value in the pipeline, and ΔH indicates the height difference between the outlet of the slurry pump and the measurement position of the pressure value in the pipeline.

[0039] Furthermore, the pipeline slurry density compensation model is implemented based on the radial basis function neural network.

[0040] Furthermore, the method further comprises:

[0041] During the real-time data collection process, the measured values ​​of pipeline slurry density are also collected regularly;

[0042] If the error between the regularly collected actual value of pipeline slurry density and the final pipeline slurry density detection result obtained based on the actual collected data at that moment is lower than the set error, the actual collected data at that moment and the actual value of pipeline slurry density are used as a set of correction data, and multiple sets of correction data form a correction data set, and the pipeline slurry density compensation model is corrected online based on the correction data set.

[0043] Furthermore, the liquid level of the slurry is detected based on a liquid level meter arranged in the slurry box.

[0044] Compared with the prior art, the present invention can achieve at least one of the following beneficial effects:

[0045] The pipeline slurry density intelligent detection method provided by the present invention has the following advantages:

[0046] First, by inputting the slurry flow value into the trained slurry pump outlet pressure prediction model, the predicted value of the slurry pump outlet pressure is obtained after processing, which is equivalent to obtaining the pressure value of a pressure point, which is equivalent to setting up a pressure gauge, thereby reducing the equipment installation and procurement costs;

[0047] Second, after obtaining the pipeline slurry density prediction value, the real-time collected slurry flow value, slurry liquid level, slurry pump motor frequency and current are input into the trained pipeline slurry density compensation model, and the pipeline slurry density error prediction value is obtained after processing; the pipeline slurry density error prediction value is used to compensate the pipeline slurry density prediction value to obtain the final pipeline slurry density detection result. This method can realize real-time and accurate measurement of slurry density, and provide guidance for the subsequent adjustment and control of slurry density in industrial sites, thereby improving production efficiency and economic benefits.

[0048] In the present invention, the above-mentioned technical solutions can also be combined with each other to achieve more preferred combination solutions. Other features and advantages of the present invention will be described in the subsequent description, and some advantages can become obvious from the description, or can be understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained through the contents particularly pointed out in the description and the drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The drawings are only for the purpose of illustrating particular embodiments and are not to be considered limiting of the present invention. Like reference symbols denote like components throughout the drawings.

[0050] Figure 1 A schematic diagram of a process flow of a pipeline slurry density intelligent detection method provided by an embodiment of the present invention;

[0051] Figure 2 A schematic diagram of the flow of a slurry pipeline provided by an embodiment of the present invention;

[0052] Figure 3 A schematic diagram of a training process of a slurry pump outlet pressure prediction model provided in an embodiment of the present invention;

[0053] Figure 4 A schematic diagram of a pipeline slurry density compensation model training process provided by an embodiment of the present invention;

[0054] Figure numerals: 1-slurry box; 2-pipeline; 3-slurry pump; 4-liquid level meter; 5-frequency meter; 6-ammeter; 7-flow meter; 8-pressure gauge. DETAILED DESCRIPTION

[0055] The preferred embodiments of the present invention are described in detail below in conjunction with the accompanying drawings, wherein the accompanying drawings constitute a part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not used to limit the scope of the present invention.

[0056] A specific embodiment of the present invention discloses a pipeline slurry density intelligent detection method, the process example is as follows Figure 1 As shown, including:

[0057] Step S1: real-time data collection, including: the pressure value in the pipeline connected to the slurry pump outlet, the slurry flow value at the slurry pump outlet, the slurry liquid level, the frequency and current of the slurry pump motor;

[0058] like Figure 2 As shown, the pressure value in the pipeline 2 connected to the slurry pump outlet is collected by the pressure gauge 8 arranged in the pipeline connected to the slurry pump 3 outlet; the slurry flow value at the slurry pump outlet is collected by the flow meter 7 arranged at the slurry pump outlet; the liquid level of the slurry is detected based on the liquid level meter 4 arranged in the slurry box 1; the frequency of the slurry pump motor is collected by the frequency meter 5 of the slurry pump motor; the current of the slurry pump motor is collected by the ammeter 6 of the slurry pump motor.

[0059] Step S2: inputting the real-time collected slurry flow value into the trained slurry pump outlet pressure prediction model, and obtaining the slurry pump outlet pressure prediction value after processing;

[0060] Step S3: obtaining a predicted value of pipeline slurry density based on the predicted value of the slurry pump outlet pressure and the pressure value in the pipeline;

[0061] Step S4: inputting the real-time collected slurry flow value, slurry liquid level, frequency and current of the slurry pump motor into the trained pipeline slurry density compensation model, and obtaining the pipeline slurry density error prediction value after processing;

[0062] Step S5: compensating the pipeline slurry density prediction value with the pipeline slurry density error prediction value to obtain a final pipeline slurry density detection result.

[0063] There is a direct relationship between the density of pipeline slurry and the pressure at different locations in the pipeline; specifically, it is reflected in:

[0064]

[0065] Among them, p1 and p2 represent the pressure values ​​of different pressure points in the pipeline connected to the outlet of the slurry pump; ΔH 1,2 Indicates the height difference between two pressure points.

[0066] According to the existing method, two pressure points need to be set in the pipeline connected to the outlet of the slurry pump. Since the pipeline material is special and expensive, drilling holes or buying a special pressure meter is relatively expensive. Therefore, reducing costs is a problem to be considered in this embodiment. In order to obtain the slurry flow parameters, some instruments are indispensable, such as the flow meter at the outlet of the slurry pump. Therefore, this embodiment obtains the corresponding pressure value based on the flow value measured by the flow meter to realize the function of a pressure meter, thereby reducing the cost of setting up two pressure meters.

[0067] First, the relationship between the relevant parameters used in this embodiment is theoretically derived and explained:

[0068] Based on the actual fluid motion law of the project, and because the slurry is transported by a slurry pump, there is energy input, so the Bernoulli equation with energy input is applied:

[0069]

[0070] Where z1 is the height difference between section 1-1 and reference plane 0-0, p1 is the relative pressure at section 1-1, ρ and g are the slurry density and gravity acceleration, respectively, a1 is the kinetic energy correction coefficient at section 1-1 (a1 = 1 when the velocity distribution is uniform), v1 is the slurry velocity at section 1-1, H m is the pump head; z2 is the height difference between section 2-2 (slurry pump outlet) and reference plane 0-0, p2 is the relative pressure at section 2-2, a2 ​​is the kinetic energy correction coefficient at section 2-2 (a2 = 1 when the flow velocity is uniformly distributed), v2 is the slurry flow velocity at section 2-2, h l is the head loss. When the properties of the transported slurry are stable, the slurry pump operates stably, and the pipeline characteristics are stable, ρ, H m and h l can be considered as a constant.

[0071] like Figure 2As shown, the ground is selected as the reference plane 0-0, the liquid level of the slurry box is the section 1-1, and the motion parameters z1 = h, p1 = 0, and v1 = 0. The horizontal plane where the flow meter at the outlet of the slurry pump is located is the section 2-2, and the motion parameter z2 is a certain value and when the liquid level z1 of the slurry box is constant, the relationship between the pressure value p2 at the section 2-2 where the flow meter is located and the slurry flow rate v2 at the section 2-2 can be obtained according to formula (2) and the above-mentioned known quantities.

[0072] The formula for calculating the slurry flow rate is: Q=v2A, where A is the cross-sectional area inside the pipeline, and the cross-sectional area of ​​the pipeline is a certain value. Therefore, there is a corresponding conversion relationship between the slurry flow value Q at section 2-2 and the pressure value p2.

[0073] like Figure 2 As shown, the ground is selected as the reference plane 0-0, the liquid level of the slurry box is the section 1-1, and the motion parameters z1=h, p1=0, and v1=0. The horizontal plane where the flow meter at the outlet of the slurry pump is located is the section 2-2, the motion parameter z2 is a certain value, and when the pipeline slurry flow Q=v2A is constant, v2 is also a certain value, and p2 is converted from the slurry flow value Q.

[0074] According to formula (2) and the above known quantities, the relationship between the liquid level height z1(h) and the slurry density ρ can be obtained. Obviously, there is a one-to-one correspondence between the liquid level height h of the slurry box and the slurry density ρ.

[0075] When the pump is driven by a motor to transport slurry in a pipeline:

[0076] Slurry pump motor power:

[0077] P=UI=P e +P c (3)

[0078] Where U and I are the voltage and current of the slurry pump motor respectively, P e is the power transmitted from the slurry pump motor to the slurry pump, P c Power lost due to other factors.

[0079] Slurry pump power:

[0080] N=N e +N c (4)

[0081] Where N e is the effective power of the slurry pump, N c The power lost due to other factors. Here, the motor power of the slurry pump is equal to the slurry pump power. e =N.

[0082] The synchronous speed of the sampling pump motor is as follows:

[0083]

[0084] Where f is the power frequency and p is the number of motor pole pairs.

[0085] Effective power of slurry pump:

[0086] N e =ρgQH e (6)

[0087] In the formula, H e Indicates the effective head of the pump.

[0088] In industrial applications, when the properties of the conveyed slurry are stable and the pump is running stably, it is generally believed that the pump flow rate Q and speed n s There is a linear correlation:

[0089] Q=μn s (7)

[0090] Among them, μ can be considered as a constant. In summary, we can get:

[0091]

[0092] When the properties of the transported slurry are stable, the slurry pump is running stably, and the pipeline characteristics are stable, μ and H e 、N c and P c are considered to be constants.

[0093] When the power supply frequency of the slurry pump motor is fixed and the voltage is stable when the motor is running, the relationship between the slurry pump motor current I and the slurry density ρ can be obtained from formula (8). Obviously, there is a one-to-one correspondence between the slurry pump motor current I and the slurry density ρ.

[0094] When the slurry pump runs stably and its motor power maintains a stable state, the relationship between the power supply frequency f and the slurry density ρ can be obtained from formula (8). Obviously, there is a one-to-one correspondence between the power supply frequency f and the slurry density ρ.

[0095] When the slurry pump runs stably and its motor power maintains a stable state, the relationship between the slurry flow rate Q and the slurry density ρ can be obtained from formula (8). Obviously, there is a one-to-one correspondence between the slurry flow rate Q and the slurry density ρ.

[0096] Compared with the traditional differential pressure density meter, the present invention introduces four process variables, namely, liquid level height h, slurry pump motor current I, power supply frequency f and slurry flow rate Q, as input variables of the pipeline slurry density compensation model through theoretical analysis, and then compensates the predicted value of pipeline slurry density. This method can better reflect the dynamic process of slurry flow, and make the slurry density detection model have better adaptability and accuracy.

[0097] Specifically, the slurry pump outlet pressure prediction model is trained in the following way. The training process is as follows: Figure 3 As shown:

[0098] Step A1: obtaining a first sample set, wherein each group of first sample data in the first sample set includes: a slurry flow value at a slurry pump outlet, a pressure value in a pipeline connected to the slurry pump outlet, and a measured value of pipeline slurry density;

[0099] Step A2: based on the measured value of pipeline slurry density and the pressure value in the pipeline in each group of first sample data, obtain the pressure standard value at the outlet of the slurry pump;

[0100] Specifically, the standard pressure value at the outlet of the slurry pump is obtained according to the following method:

[0101] p er (i) = ρ er (i)gΔH+p c (i) (9)

[0102] Among them, ρ er (i) represents the measured value of pipeline slurry density in the first sample data of the i-th group, p c (i) represents the pressure value in the pipeline connected to the slurry pump outlet in the i-th group of first sample data, and ΔH represents the height difference between the slurry pump outlet and the measurement position of the pressure value in the pipeline.

[0103] Step A3: Take the slurry flow value at the slurry pump outlet in each group of first sample data as input and the pressure standard value at the slurry pump outlet as a label, train the slurry pump outlet pressure prediction model, determine the structure and parameters of the slurry pump outlet pressure prediction model, and obtain a trained slurry pump outlet pressure prediction model.

[0104] Specifically, the structure and parameters of the slurry pump outlet pressure prediction model are determined based on the LS least squares method. Preferably, the structure and parameters of the slurry pump outlet pressure prediction model are determined in the following manner:

[0105] Predicted pressure value p at the slurry pump outlet ef The slurry flow rate Q at the outlet of the slurry pump satisfies:

[0106] p ef =aQ 2 +b (10)

[0107] The standard pressure value and predicted pressure value at the slurry pump outlet meet the following requirements:

[0108]

[0109] Where n represents the number of groups of the first sample data in the first sample set; p er (i) represents the standard pressure value at the outlet of the slurry pump in the first sample data of the i-th group, p ef (i) represents the predicted pressure value at the outlet of the slurry pump in the first sample data of the i-th group;

[0110] Based on the LS least squares method, the minimum value of J(a,b) is taken as the goal, and the parameters a and b are estimated to obtain the estimated value a of parameter a. * , the estimated value of parameter b * ;

[0111] The trained slurry pump outlet pressure prediction model is:

[0112] p ef =a * Q 2 +b * (12).

[0113] In this embodiment, the pipeline slurry density compensation model is trained in the following manner. The training process is as follows: Figure 4 As shown:

[0114] Step B1: obtaining a second sample set, wherein each set of second sample data in the second sample set includes: a measured value of pipeline slurry density, a pressure value in a pipeline connected to a slurry pump outlet, a slurry flow value at a slurry pump outlet, a slurry liquid level, and a frequency and current of a slurry pump motor;

[0115] Step B2: inputting the slurry flow value at the slurry pump outlet in each set of second sample data into the trained slurry pump outlet pressure prediction model, and obtaining the corresponding slurry pump outlet pressure prediction value after processing;

[0116] Step B3: based on the predicted value of the slurry pump outlet pressure and the pressure value in the pipeline corresponding to each set of second sample data, obtain the corresponding predicted value of the pipeline slurry density;

[0117] Specifically, the pipeline slurry density prediction value ρ is calculated according to the following formula: f ,include:

[0118]

[0119] Among them, p ef represents the predicted pressure at the outlet of the slurry pump, p c It indicates the pressure value in the pipeline, and ΔH indicates the height difference between the outlet of the slurry pump and the measurement position of the pressure value in the pipeline.

[0120] Step B4: Take the slurry flow value at the slurry pump outlet, the liquid level of the slurry, the frequency and current of the slurry pump motor in each group of second sample data as input, and the difference between the pipeline slurry density prediction value and the pipeline slurry density actual value as a label, train the pipeline slurry density compensation model, and obtain a trained pipeline slurry density compensation model.

[0121] Preferably, the pipeline slurry density compensation model is implemented based on a radial basis function neural network.

[0122] In addition, the method further comprises:

[0123] During the real-time data collection process, the measured values ​​of pipeline slurry density are also collected regularly;

[0124] If the error between the regularly collected actual value of pipeline slurry density and the final pipeline slurry density detection result obtained based on the actual collected data at that moment is lower than the set error, the actual collected data at that moment and the actual value of pipeline slurry density are used as a set of correction data, and multiple sets of correction data form a correction data set. The pipeline slurry density compensation model is corrected online based on the correction data set, and the online correction process refers to the training process of the model.

[0125] In summary, compared with the prior art, the pipeline slurry density intelligent detection method provided by the embodiment of the present invention has the following advantages: first, by inputting the slurry flow value into the trained slurry pump outlet pressure prediction model, the slurry pump outlet pressure prediction value is obtained after processing, which is equivalent to obtaining the pressure value of a pressure point, which is equivalent to setting a pressure gauge, thereby reducing the equipment installation and procurement costs; second, after obtaining the rear pipeline slurry density prediction value, the real-time collected slurry flow value, the liquid level of the slurry, the frequency and current of the slurry pump motor are also input into the trained pipeline slurry density compensation model, and the pipeline slurry density error prediction value is obtained after processing; the pipeline slurry density error prediction value is used to compensate the pipeline slurry density prediction value to obtain the final pipeline slurry density detection result. This method can realize real-time and accurate measurement of slurry density, and provide guidance for the subsequent adjustment and control of slurry density in industrial sites, thereby improving production efficiency and economic benefits.

[0126] Those skilled in the art will appreciate that all or part of the processes of the above-mentioned embodiments can be implemented by instructing related hardware through a computer program, and the program can be stored in a computer-readable storage medium, wherein the computer-readable storage medium is a disk, an optical disk, a read-only storage memory, or a random access memory, etc.

[0127] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by any technician familiar with the technical field within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention.

Claims

1. A pipeline slurry density intelligent detection method, characterized in that: include: Real-time data collection, including: the pressure value in the pipeline connected to the slurry pump outlet, the slurry flow value at the slurry pump outlet, the slurry liquid level, the frequency and current of the slurry pump motor; The real-time collected slurry flow value is input into the trained slurry pump outlet pressure prediction model, and the slurry pump outlet pressure prediction value is obtained after processing; Based on the predicted value of the slurry pump outlet pressure and the pressure value in the pipeline, the predicted value of the pipeline slurry density is obtained; The real-time collected slurry flow value, slurry liquid level, frequency and current of the slurry pump motor are input into the trained pipeline slurry density compensation model, and the pipeline slurry density error prediction value is obtained after processing; the slurry is transported from the slurry box to the pipeline connected to the slurry pump outlet through the slurry pump, and the slurry liquid level is detected based on the liquid level meter set in the slurry box; The pipeline slurry density prediction value is compensated by the pipeline slurry density error prediction value to obtain a final pipeline slurry density detection result; The slurry pump outlet pressure prediction model is trained in the following way: Acquire a first sample set, wherein each set of first sample data in the first sample set includes: a slurry flow value at a slurry pump outlet, a pressure value in a pipeline connected to the slurry pump outlet, and a measured value of pipeline slurry density; Based on the measured value of pipeline slurry density and the pressure value in the pipeline in each group of first sample data, the standard value of pressure at the outlet of the slurry pump is obtained; Taking the slurry flow value at the outlet of the slurry pump in each group of first sample data as input and the pressure standard value at the outlet of the slurry pump as a label, training the slurry pump outlet pressure prediction model, determining the structure and parameters of the slurry pump outlet pressure prediction model, and obtaining a trained slurry pump outlet pressure prediction model; The pipeline slurry density compensation model is trained in the following way: Acquire a second sample set, wherein each set of second sample data in the second sample set includes: a measured value of pipeline slurry density, a pressure value in a pipeline connected to a slurry pump outlet, a slurry flow value at a slurry pump outlet, a liquid level of slurry, and a frequency and current of a slurry pump motor; Inputting the slurry flow rate value at the slurry pump outlet in each group of second sample data into the trained slurry pump outlet pressure prediction model, and obtaining the corresponding slurry pump outlet pressure prediction value after processing; Based on the predicted value of the slurry pump outlet pressure and the pressure value in the pipeline corresponding to each set of second sample data, the corresponding predicted value of the pipeline slurry density is obtained; The slurry flow value at the slurry pump outlet, the liquid level of the slurry, the frequency and current of the slurry pump motor in each group of second sample data are used as input, and the difference between the pipeline slurry density prediction value and the pipeline slurry density actual value is used as a label to train the pipeline slurry density compensation model to obtain a trained pipeline slurry density compensation model.

2. The intelligent detection method for pipeline slurry density according to claim 1 is characterized in that: In the slurry pump outlet pressure prediction model, the structure and parameters of the slurry pump outlet pressure prediction model are determined based on the LS least squares method.

3. The intelligent detection method for pipeline slurry density according to claim 2 is characterized in that: The structure and parameters of the slurry pump outlet pressure prediction model are determined by the following method: Predicted pressure at the slurry pump outlet and the slurry flow value at the outlet of the slurry pump satisfy: （1） The standard pressure value and predicted pressure value at the slurry pump outlet meet the following requirements: （2） in, represents the number of groups of first sample data in the first sample set; Indicates The standard pressure value at the outlet of the slurry pump in the first sample data of the group, Indicates The predicted pressure value at the outlet of the slurry pump in the first sample data of the group; Based on the LS least squares method, Take the minimum value as the target, and and Estimate and obtain the parameters Estimated value of ,parameter Estimated value of ; The trained slurry pump outlet pressure prediction model is: （3）。 4. The intelligent detection method for pipeline slurry density according to claim 3 is characterized in that: The standard pressure value at the outlet of the slurry pump can be obtained by the following method: （4） in, Indicates The measured value of pipeline slurry density in the first sample data of group 1, Indicates The pressure value in the pipeline connected to the slurry pump outlet in the first sample data group, Indicates the height difference between the slurry pump outlet and the measurement location of the pressure value in the pipeline.

5. The intelligent detection method for pipeline slurry density according to claim 1 is characterized in that: The predicted value of pipeline slurry density is calculated according to the following formula ,include: （5） in, Indicates the predicted pressure value at the outlet of the slurry pump, Indicates the pressure value in the pipeline. Indicates the height difference between the slurry pump outlet and the measurement location of the pressure value in the pipeline.

6. The intelligent detection method for pipeline slurry density according to claim 5 is characterized in that: The pipeline slurry density compensation model is implemented based on radial basis function neural network.

7. The intelligent detection method for pipeline slurry density according to any one of claims 1 to 6, characterized in that: The method further comprises: During the real-time data collection process, the measured values ​​of pipeline slurry density are also collected regularly; If the error between the regularly collected actual value of pipeline slurry density and the final pipeline slurry density detection result obtained based on the actual collected data at that moment is lower than the set error, the actual collected data at that moment and the actual value of pipeline slurry density are used as a set of correction data, and multiple sets of correction data form a correction data set, and the pipeline slurry density compensation model is corrected online based on the correction data set.

Images