A feeding method for coal slime water sedimentation
By calculating and optimizing the proportional coefficients in the PID algorithm, using prediction models and neural networks to optimize the loading flow rate, the problem of poor loading flow rate adjustment in the existing technology is solved, and the settlement efficiency and settlement quality of coal sludge water are improved.
Patent Information
- Application Number
- CN202411284832.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-13
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2044-09-13
AI Technical Summary
When the prior art uses the PID algorithm to adjust the feeding flow rate of coal sludge water, the proportional coefficient is too small and leads to weak control effect. If it is too large, it may cause oscillation, affecting the sedimentation efficiency of coal sludge water.
By calculating the optimal proportional coefficient in the PID algorithm, the prediction model and neural network are used to optimize the loading flow rate, so that it can achieve the optimal flow rate in the shortest time, reduce regulation oscillation, and optimize the settlement effect.
The loading flow rate is adjusted to the optimal state in the shortest time, reducing oscillation during the adjustment process, and improving the settlement efficiency and settlement quality of coal sludge water.
Smart Images

Figure CN118791068B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of coal slime water sedimentation, and in particular to a feeding method for coal slime water sedimentation. Background Art
[0002] Coal slurry sedimentation refers to the process in which solid particles such as coal dust and slag contained in wastewater settle to the bottom after a certain period of time during coal mining, thereby separating water from sediments. In the process of coal slurry sedimentation, the generated coal slurry water is usually discharged into a sedimentation tank for sedimentation. The process of discharging coal slurry water into a sedimentation tank is called feeding. Feeding plays a vital role in the sedimentation effect. If the flow rate of feeding is too large, it is easy to impact the sediments that have already settled, affecting the sedimentation effect. If the flow rate of feeding is too small, the discharge time of wastewater is extended and the sedimentation efficiency is reduced. By accurately controlling the flow rate of feeding, the sedimentation efficiency can be improved while ensuring the sedimentation effect, and the solid particles and water can be effectively separated, thereby achieving effective wastewater treatment and resource recovery.
[0003] The Chinese patent application document with publication number CN115852345A discloses a temperature control method for ALD equipment based on integral separation PID algorithm. It includes steps S1, obtaining the objective function of the system according to the temperature rise curve of the ALD equipment; S2, using the Simlink module in MATLAB as a tool, analyzing the control algorithm of the objective function and the integral term switching coefficient of the integral separation PID algorithm, and adjusting the parameters of the PID algorithm; S3, using the temperature control method for ALD equipment based on the integral separation PID algorithm to control the temperature.
[0004] The PID algorithm is a control algorithm that combines the three links of proportion, integration and differentiation. In the above scheme, the integral term is adjusted. However, when using the PID algorithm to adjust the feed flow rate,
[0005] When the proportional coefficient is too small, the feeding flow rate control effect is weak, and the transition to the optimal flow rate is slow; when the proportional coefficient is too large, the possibility of oscillation is high and the system stability is poor, both of which will lead to poor feeding flow rate and reduced coal slurry water sedimentation efficiency.
[0006] How to choose the best proportional coefficient to control the feeding flow rate of coal slurry water so as to achieve the best sedimentation effect of slurry water is a technical problem that needs to be solved at present. Summary of the invention
[0007] In order to solve the problem of how to select the best proportional coefficient to control the feeding flow rate of coal slime water, the present invention provides a feeding method for coal slime water sedimentation.
[0008] A feeding method for coal slime water sedimentation, comprising the steps of:
[0009] Calculate the optimal proportional coefficient in the PID algorithm, and use the PID algorithm to adjust the feeding flow rate at the future moment to achieve the optimal feeding flow rate;
[0010] Among them, the calculation method of the optimal proportional coefficient is:
[0011] Set the initial proportionality coefficient, obtain the adjustment period corresponding to the initial proportionality coefficient and the feeding flow rate at each moment in the adjustment period;
[0012] According to the feed flow rate at each moment, the settling mass at the future moment is obtained by using the preset prediction model, and the difference of the settling mass at two adjacent moments is taken as the convergence efficiency;
[0013] Calculate the optimization degree of the proportional coefficient, which is negatively correlated with the absolute value of the convergence efficiency and the sum of the adjustment time;
[0014] The initial proportional coefficient is iteratively updated to obtain multiple proportional coefficients, and the optimization degree of each proportional coefficient is calculated respectively, and the proportional coefficient with the largest optimization degree is taken as the optimal proportional coefficient of the PID algorithm at the future moment.
[0015] The priority of different proportional coefficients is calculated, the optimal proportional coefficient is obtained according to the priority, and the feeding flow rate at the next moment is adjusted according to the optimal proportional coefficient. The feeding flow rate can be adjusted to the optimal feeding flow rate in the shortest time, and the oscillation interval during the adjustment process is short, the adjustment oscillation is faster, and the sedimentation effect of the coal slurry water is improved.
[0016] Preferably, the expression of sedimentation mass is:
[0017]
[0018] In the formula, For the The sedimentation mass of coal slime water at each moment; For the Separation efficiency of the supernatant at each moment; For the Sedimentation efficiency at each moment; () is a linear normalization function.
[0019] The formula can accurately calculate the sedimentation mass of coal slime water, which is convenient for training the prediction model.
[0020] Preferably, the preferred degree of the proportionality coefficient is expressed as:
[0021]
[0022] In the formula, Indicates The preference of the proportionality coefficients; Indicates The PID algorithm corresponding to the proportional coefficient makes the feed flow rate converge to the adjustment time; express Momentary settling mass convergence efficiency; Indicates the The PID algorithm corresponding to the proportional coefficient makes the moment when the convergence efficiency is less than 0 for the first time when the feeding flow rate converges; is a linear normalization function.
[0023] Preferably, the preferred degree of the proportionality coefficient is expressed as:
[0024]
[0025] In the formula, Indicates The preference of the proportionality coefficients; Indicates The PID algorithm corresponding to the proportional coefficient makes the feed flow rate converge to the adjustment time; express Momentary settling mass convergence efficiency; Indicates the The PID algorithm corresponding to the proportional coefficient makes the moment when the convergence efficiency is less than 0 for the first time when the feeding flow rate converges; is a linear normalization function.
[0026] The formula can accurately calculate the optimization degree of the proportional coefficient, which is convenient for selecting the best proportional coefficient.
[0027] Preferably, the prediction model is obtained by:
[0028] A neural network model is constructed to obtain a training set, wherein the training set includes multiple groups of parameter sequences, each group of parameter sequences is set with a corresponding label, and the label is the sedimentation mass;
[0029] The training set is input into the pre-built neural network model, and the mean square error loss function between the output sedimentation mass and the label is calculated;
[0030] The prediction model parameters are adjusted using the gradient descent method to minimize the error between the output value and the label, and the prediction model parameters are adjusted iteratively until the mean square error loss function is less than the set value or the preset number of iterations is reached to obtain a trained prediction model.
[0031] The prediction model can predict the sedimentation mass corresponding to each parameter sequence at the next moment, so as to select the optimal feeding flow rate at the next moment and achieve the optimal sedimentation effect of coal slurry water.
[0032] Preferably, the parameter sequence includes the clarity of the supernatant of the coal slime water, the height of the sediment, value, temperature, concentration and feed flow rate.
[0033] The clarity of the supernatant of coal sludge water, the height of sediment, Value, temperature and concentration have a great influence on the sedimentation of coal slime water. By selecting relevant parameters of coal slime water, it is convenient to accurately predict the sedimentation quality in the subsequent process.
[0034] Preferably, the method of predicting the sedimentation mass using a prediction model according to the feed flow rate at each moment is:
[0035] Get the clarity of the supernatant liquid, the height of the sediment, value, temperature, and concentration; obtain the feed flow rate at each moment in the adjustment process, and then construct a parameter sequence, with each moment corresponding to a parameter sequence;
[0036] The parameter sequence at each moment is input into the prediction model to obtain the sedimentation mass at each moment.
[0037] Preferably, the method for obtaining the optimal feeding flow rate is:
[0038] Get the clarity of the supernatant liquid, the height of the sediment, value, temperature, concentration, set the initial feed flow rate to build a parameter sequence;
[0039] Iteratively updating the feed flow rate in the parameter sequence to obtain multiple parameter sequences;
[0040] Inputting multiple parameter sequences into the prediction network model in sequence to obtain multiple sedimentation masses;
[0041] The feeding flow rate in the parameter sequence corresponding to the maximum value of the sedimentation mass is taken as the optimal feeding flow rate at the future moment.
[0042] The present invention has the following technical effects:
[0043] The optimal feeding flow rate is selected according to the prediction model, and the PID algorithm is used to take the optimal feeding flow rate as the control target. Different proportional coefficients are iteratively selected, and then the priority of different proportional coefficients is calculated. The optimal proportional coefficient is obtained according to the priority, and the PID algorithm corresponding to the optimal proportional coefficient is used to adjust the feeding flow rate at the next moment. The feeding flow rate can be adjusted to the optimal feeding flow rate in the shortest time, and the oscillation time during the adjustment process is longer, which improves the sedimentation effect of the coal slime water. BRIEF DESCRIPTION OF THE DRAWINGS
[0044] By reading the detailed description below with reference to the accompanying drawings, the above and other purposes, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-restrictive manner, and the same or corresponding reference numerals represent the same or corresponding parts.
[0045] Figure 1 The present invention is a method flow chart of a feeding method for coal slime water sedimentation.
[0046] Figure 2 It is a comparison diagram of the effects of the best proportional coefficient and other proportional coefficients in a feeding method for coal slime water sedimentation of the present invention. DETAILED DESCRIPTION
[0047] 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 part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.
[0048] It should be understood that when the terms "first", "second", etc. are used in the claims, descriptions, and drawings of the present invention, they are only used to distinguish different objects, rather than to describe a specific order. The terms "include" and "comprise" used in the description and claims of the present invention indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their collections.
[0049] The embodiment of the present invention discloses a feeding method for coal slurry water sedimentation, which solves the problem of how to select the best proportional coefficient to control the coal slurry water feeding flow rate. Figure 1 , including steps S1 to S3, which are specifically as follows:
[0050] S1. Obtain multiple parameter sequences of coal slime water at the current moment, and input the multiple parameter sequences into the prediction model to obtain the sedimentation mass corresponding to each parameter sequence, wherein the feed flow rates in the parameter sequences are all different.
[0051] S101, parameter sequence includes supernatant clarity of coal slurry, sediment height, value, temperature, concentration and feed flow rate.
[0052] A turbidity meter is arranged at the outlet of the supernatant liquid of the coal slurry water settling tank to collect the turbidity of the supernatant liquid during the sedimentation process of the coal slurry water, and the reciprocal of the measured supernatant liquid turbidity is used as the clarity of the supernatant liquid; an industrial camera is arranged on the side of the coal slurry water settling tank to collect images of the settling tank and obtain the sediment height in the image through image recognition. Measuring instrument measures the coal sludge water in the sedimentation tank The temperature and concentration of the coal slurry water are measured by installing a temperature sensor and a liquid concentration meter in the sedimentation tank.
[0053] For example, the parameter sequence is ( , , …, ),in, , , …, They correspond to the clarity of the supernatant, the height of the sediment, value, temperature, concentration, is the feeding flow rate, iteratively updated The value of , , , then the obtained parameter sequence is ( , , …, )、( , , …, )、( , , …, )、( , , …, ).
[0054] S102: training a prediction model.
[0055] A training set is obtained, wherein the training set includes multiple groups of parameter sequences, each group of environmental data is set with a corresponding label, and the label is the sedimentation mass; the training set is input into a pre-built neural network model for training; the mean square error loss function between the output sedimentation mass and the label is calculated; the prediction model parameters are adjusted using a gradient descent method to minimize the error between the output value and the label; the parameters of the prediction model are iteratively adjusted until the mean square error loss function is less than a set value or a preset number of iterations is reached, thereby obtaining a trained prediction model; multiple parameter sequences are respectively input into the prediction model to obtain the corresponding sedimentation mass, and the neural network model adopts a recurrent neural network.
[0056] The expression of sedimentation mass is:
[0057]
[0058] In the formula, For the The sedimentation mass of coal slime water at each moment; For the Separation efficiency of the supernatant at each moment; For the Sedimentation efficiency at each moment; () is a linear normalized function. According to the expression of sedimentation mass, corresponding labels can be set for each set of environmental data. The larger the sedimentation mass value, the better the sedimentation effect of the coal slime water, the faster the separation of coal slime water recovery water, and the higher the sedimentation efficiency, which is conducive to the subsequent treatment of coal slime water.
[0059] In this embodiment, is the difference between the clarity of the supernatant at the current moment and the previous moment, is the difference between the sediment height at the current moment and the previous moment.
[0060] In other embodiments, is the ratio of the clarity of the supernatant at the current moment to that at the previous moment, It is the ratio of the sediment height at the current moment to the previous moment.
[0061] S103, a method of inputting a plurality of parameter sequences into the prediction model to obtain the sedimentation mass corresponding to each parameter sequence at the next moment is:
[0062] For example, get the current parameters , , …, , artificially set the feeding flow rate , forming a parameter sequence ( , , …, ),Change The value of , , …, )、( , , …, )、( , , …, ), the parameter sequence ( , , …, )、( , , …, )、( , , …, )、( , , …, ), which are input into the recurrent neural network in sequence to obtain four sedimentation masses.
[0063] S2. The feeding flow rate in the parameter sequence with the largest sedimentation mass value is taken as the optimal feeding flow rate at the next moment.
[0064] Exemplarily, in step S102, the parameter sequence ( , ,……, ) corresponds to the largest sedimentation mass value, then select It is the optimal feeding flow rate at the next moment.
[0065] S3. Use PID algorithm to adjust the feeding flow rate at the next moment to achieve the optimal feeding flow rate.
[0066] The method of using PID algorithm to adjust the feeding flow rate at the next moment is:
[0067] S301, setting an initial proportional coefficient, obtaining an adjustment period corresponding to the initial proportional coefficient and a material feeding flow rate at each moment in the adjustment period, iteratively updating the initial proportional coefficient and obtaining an adjustment period corresponding to each proportional coefficient and a material feeding flow rate at each moment in the adjustment period.
[0068] Among them, the adjustment period is the adjustment time required to adjust to the optimal feeding flow rate. The optimal feeding flow rate is taken as the adjustment target, and the feeding flow rate is adjusted using the PID algorithm, so that the feeding flow rate has less oscillation during the adjustment process and the adjustment time is less.
[0069] For example, set the initial scale factor , integral coefficient , differential coefficient It should be noted that the integral coefficient and differential coefficient are set manually according to experience, and the proportional coefficient is selected iteratively. , , …, , when the selected proportional coefficient produces oscillation, the iteration stops.
[0070] The optimal feed flow rate is used as the target value for PID algorithm adjustment. Proportional Factor Under the regulation, the PID algorithm is used to obtain the regulation time required for the regulation process and the feeding flow rate at each moment. This is the existing technology in the PID algorithm, and the specific process will not be repeated here.
[0071] S302, predicting the sedimentation mass using a prediction model according to the feed flow rate at each moment, and calculating the convergence efficiency at each moment, where the convergence efficiency is the difference between the sedimentation masses at two adjacent moments.
[0072] Get the supernatant clarity, sediment height, value, temperature, and concentration; the feed flow rate at each moment is added to the parameter sequence to obtain multiple updated parameter sequences, and the updated parameter sequence includes supernatant clarity, sediment height, Value, temperature, concentration, feed flow rate. The updated multiple parameter sequences are sequentially input into the prediction model to obtain the corresponding sedimentation mass.
[0073] In the Proportional Factor In the corresponding PID algorithm, during the adjustment process Coal slime water sedimentation quality at all times , No. Coal slime water sedimentation quality at each moment , then Momentary sedimentation mass convergence efficiency for: , where is a normalized function. The larger its value is, the better the sedimentation quality of the coal slurry water is and the faster the sedimentation speed is.
[0074] S303, calculate the preference of the corresponding proportional coefficient, the preference is negatively correlated with the absolute value of the convergence efficiency less than 0 and the sum of the adjustment time, and take the proportional coefficient with the largest preference as the optimal proportional coefficient of the PID algorithm at the next moment.
[0075] In one embodiment, the preferred degree of the proportionality coefficient is expressed as:
[0076]
[0077] In the formula, Indicates The preference of the proportionality coefficients; Indicates The PID algorithm corresponding to the proportional coefficient makes the feed flow rate converge to the adjustment time; express Momentary settling mass convergence efficiency; Indicates the The PID algorithm corresponding to the proportional coefficient makes the moment when the convergence efficiency is less than 0 for the first time when the feeding flow rate converges; is a linear normalized function. If the convergence efficiency is not less than 0, .
[0078] It should be noted that in Proportional Factor In the corresponding PID algorithm, after the initial proportional coefficient, integral coefficient, and differential coefficient are determined, the adjustment time for the feed flow rate to converge can be calculated. This is the prior art, and the specific steps will not be repeated here. This is the value at which the sedimentation mass convergence efficiency first becomes negative.
[0079] In other embodiments, the preferred expression of the proportionality coefficient is:
[0080]
[0081] In the formula, Indicates The preference of the proportionality coefficients; Indicates The PID algorithm corresponding to the proportional coefficient makes the feed flow rate converge to the adjustment time; express Momentary settling mass convergence efficiency; Indicates the The PID algorithm corresponding to the proportional coefficient makes the moment when the convergence efficiency is less than 0 for the first time when the feeding flow rate converges; is a linear normalized function. If the convergence efficiency is not less than 0, .
[0082] Get The first time the convergence efficiency is less than 0 is when the proportional coefficient is adjusted. Linear normalized absolute sedimentation mass convergence efficiency and linear normalized adjustment interval The smaller the sum is, the shorter the oscillation interval is and the faster the sedimentation mass converges. The shorter the oscillation interval is, the faster the subsequent adjustment oscillation is. The higher the total regulation efficiency for adjusting the proportional coefficient, the higher the proportional coefficient The higher the preference.
[0083] The optimality of each proportional coefficient is calculated according to the formula, and the proportional coefficient with the largest optimality is selected as the proportional coefficient at the next moment to adjust the feeding flow rate. During the adjustment process, the feeding flow rate can reach the optimal feeding flow rate in the shortest time, and the oscillation phenomenon in the adjustment process is reduced, making the adjustment process more stable.
[0084] like Figure 2 As shown, the horizontal axis is time and the vertical axis is the feeding flow rate. In the process of adjusting the feeding flow rate, the PID algorithm with the optimal proportional coefficient takes less time to reach the target feeding flow rate and has a shorter oscillation time than the PID algorithms with other proportional coefficients. It can ensure the sedimentation quality of the coal slurry water and achieve the optimal sedimentation effect.
[0085] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.
[0086] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.
Claims
1. A feeding method for coal slime water sedimentation, characterized in that: Includes steps: Calculate the optimal proportional coefficient in the PID algorithm, and use the PID algorithm to adjust the feeding flow rate at the future moment to achieve the optimal feeding flow rate; The calculation method of the optimal proportional coefficient is: Set the initial proportionality coefficient, obtain the adjustment period corresponding to the initial proportionality coefficient and the feeding flow rate at each moment in the adjustment period; According to the feed flow rate at each moment, the preset prediction model is used to obtain the sedimentation mass at the future moment. The expression is: In the formula, For the The sedimentation mass of coal slime water at each moment; For the The separation efficiency of the supernatant at each moment; For the Sedimentation efficiency at each moment; () is the linear normalization function; The difference of the sedimentation mass at two adjacent moments is taken as the convergence efficiency; Calculate the optimality of the proportionality coefficient, expressed as: In the formula, Indicates The preference of the proportionality coefficients; Indicates The PID algorithm corresponding to the proportional coefficient makes the feed flow rate converge to the adjustment time; express Momentary settling mass convergence efficiency; Indicates the The PID algorithm corresponding to the proportional coefficient makes the moment when the convergence efficiency is less than 0 for the first time when the feeding flow rate converges; It is a linear normalized function, and the optimization degree is negatively correlated with the absolute value of the convergence efficiency and the sum of the adjustment time; The initial proportional coefficient is iteratively updated to obtain multiple proportional coefficients, and the optimization degree of each proportional coefficient is calculated respectively, and the proportional coefficient with the largest optimization degree is taken as the optimal proportional coefficient of the PID algorithm at the future moment.
2. A feeding method for coal slime water sedimentation according to claim 1, characterized in that: The preferred expression of the proportionality coefficient is: In the formula, Indicates The preference of the proportionality coefficients; Indicates The PID algorithm corresponding to the proportional coefficient makes the feed flow rate converge to the adjustment time; express Momentary settling mass convergence efficiency; Indicates the The PID algorithm corresponding to the proportional coefficient makes the moment when the convergence efficiency is less than 0 for the first time when the feeding flow rate converges; is a linear normalization function.
3. A feeding method for coal slime water sedimentation according to claim 1, characterized in that: The prediction model is obtained as follows: Build a neural network model; Obtaining a training set, wherein the training set includes multiple groups of parameter sequences, each group of parameter sequences is set with a corresponding label, and the label is sedimentation mass; The training set is input into the pre-built neural network model, and the mean square error loss function between the output sedimentation mass and the label is calculated; The prediction model parameters are adjusted using the gradient descent method to minimize the error between the output value and the label, and the prediction model parameters are iteratively adjusted until the mean square error loss function is less than the set value or the preset number of iterations is reached to obtain a trained prediction model.
4. A feeding method for coal slime water sedimentation according to claim 3, characterized in that: The parameter sequence includes the clarity of the supernatant of the coal slurry, the height of the sediment, value, temperature, concentration and feed flow rate.
5. A feeding method for coal slime water sedimentation according to claim 4, characterized in that: The method of using the prediction model to predict the sedimentation mass according to the feed flow rate at each moment is: Get the clarity of the supernatant liquid, the height of the sediment, value, temperature, and concentration; obtain the feed flow rate at each moment in the adjustment process, and then construct a parameter sequence, with each moment corresponding to a parameter sequence; The parameter sequence at each moment is input into the prediction model to obtain the sedimentation mass at each moment.
6. A feeding method for coal slime water sedimentation according to claim 4, characterized in that: The method to obtain the optimal feeding flow rate is: Get the clarity of the supernatant liquid, the height of the sediment, value, temperature, concentration, set the initial feed flow rate to build a parameter sequence; Iteratively updating the feed flow rate in the parameter sequence to obtain multiple parameter sequences; Inputting multiple parameter sequences into the prediction network model in sequence to obtain multiple sedimentation masses; The feeding flow rate in the parameter sequence corresponding to the maximum value of the sedimentation mass is taken as the optimal feeding flow rate at the future moment.
