CN112363391A - Sludge bulking inhibition method based on self-adaptive segmented sliding mode control - Google Patents
Sludge bulking inhibition method based on self-adaptive segmented sliding mode control Download PDFInfo
- Publication number
- CN112363391A CN112363391A CN202011092548.5A CN202011092548A CN112363391A CN 112363391 A CN112363391 A CN 112363391A CN 202011092548 A CN202011092548 A CN 202011092548A CN 112363391 A CN112363391 A CN 112363391A
- Authority
- CN
- China
- Prior art keywords
- time
- sludge
- concentration
- neuron
- sliding mode
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Feedback Control In General (AREA)
- Activated Sludge Processes (AREA)
Abstract
The invention provides a sludge bulking restraining method based on self-adaptive segmented sliding mode control, which aims at solving the problems that the abnormal working condition phenomenon of sludge bulking is easy to occur in the sewage treatment process and the transition to the normal working condition is difficult, and realizes the restraint of sludge bulking. According to the invention, the fuzzy neural network is adopted to obtain the water quality parameters of the sewage treatment operation process, the sludge volume index is predicted in real time, the working condition of the operation process is judged, a reference signal is provided for switching of the controller, and the segmented sliding mode controller with the self-adaptive switching mechanism is designed to regulate and control the dissolved oxygen concentration and the nitrate nitrogen concentration, coordinate the biochemical reaction process, improve the sludge sedimentation characteristic, and ensure that the sewage treatment process is recovered to the normal working condition when sludge expansion occurs.
Description
Technical Field
The invention utilizes a sludge bulking restraining method based on self-adaptive segmented sliding mode control to realize the restraining of sludge bulking in the sewage treatment process, and coordinates the biochemical reaction process by regulating and controlling the process variables of dissolved oxygen concentration and nitrate nitrogen concentration to stabilize the sewage treatment process or reach the normal working condition again; the sludge volume index is an important index for measuring sludge bulking, and the real-time prediction of the index in the control process has important significance; the method ensures that the sewage treatment process recovers to a normal working condition when sludge bulking occurs, and belongs to the field of water research and intelligent control;
background
Water is the most important basic resource for urban development. In recent years, the continuous development of industry causes the sewage discharge to be gradually increased, thereby restricting national economy and influencing ecological environment. The urban sewage treatment process can realize sustainable utilization and virtuous cycle of water resources, and is an important measure for urban development. The process has continuity and irreplaceability of work, and once abnormal working conditions occur, the operation of the whole treatment process is influenced, and huge economic loss and important social influence are generated. Sludge bulking is one of the most typical abnormal conditions and is a major bottleneck limiting the normal operation of a sewage treatment process.
The harm of sludge bulking is mainly reflected in two aspects: slight sludge expansion can cause insufficient sludge compression in the secondary sedimentation tank, the backflow efficiency is low, and the sewage treatment capacity is reduced; severe sludge bulking can damage the secondary sedimentation tank, causing sludge-water separation and ultimately leading to a breakdown of the entire system. The sludge bulking mechanism is complex, the complex microbial growth process and various related factors are involved, and great challenges are brought to inhibiting sludge bulking and maintaining stable operation of the sewage treatment process. At present, the sludge volume index is the most common standard for determining the sludge bulking grade and can be used for measuring the sedimentation performance of activated sludge. The sludge volume index is generally determined by an off-line analysis technique in a laboratory, and is difficult to automatically monitor and control. Therefore, inhibiting sludge expansion and ensuring reliable and stable operation of the sewage treatment process are problems to be solved urgently; the sludge bulking suppression method based on the self-adaptive segmented sliding mode control can adjust the process variable dissolved oxygen concentration and nitrate nitrogen concentration, predict the sludge volume index in real time, coordinate the biochemical reaction process and improve the sludge sedimentation characteristic to realize the suppression of the sludge bulking in the sewage treatment process, so that the sewage treatment process is kept in a normal working condition, and the method has important significance for the sustainable utilization of water resources and the environmental protection.
The invention designs a sludge bulking restraining method based on self-adaptive segmented sliding mode control, which mainly constructs a segmented sliding mode controller and designs a self-adaptive switching mechanism by predicting the sludge volume index in the sewage treatment process in real time to complete the restraining of sludge bulking.
Disclosure of Invention
The invention obtains a sludge bulking inhibition method based on self-adaptive segmented slip-form control, which predicts the sludge volume index of a sewage treatment process in real time by utilizing the dissolved oxygen concentration of an aerobic tank, the sludge load rate of a secondary sedimentation tank, the suspended solid concentration of mixed liquid of the secondary sedimentation tank, the retention time and the reflux ratio of sludge of the secondary sedimentation tank, constructs a multi-segment slip-form controller to adjust the dissolved oxygen concentration and the nitrate nitrogen concentration, and designs a self-adaptive switching mechanism to keep the sewage treatment process in a normal working condition and realize the inhibition of sludge bulking;
the invention adopts the following technical scheme and implementation steps:
1. a sludge bulking restraining method based on self-adaptive sectional sliding mode control,
the sludge bulking is restrained by predicting the sludge volume index in real time and designing a proper sliding mode controller;
the method is characterized by comprising the following steps:
(1) a fuzzy neural network for real-time prediction of sludge volume index is designed, and the structure of the fuzzy neural network is divided into four layers:
the device comprises an input layer, a radial base layer, a normalization layer and an output layer; the method specifically comprises the following steps:
an input layer: this layer includes 5 input neurons:
α(p)＝[α1(p),α2(p),α3(p),α4(p),α5(p)]T (1)
where α (p) is the input vector of the fuzzy neural network at time p, α1(p) is the dissolved oxygen concentration of the aerobic tank at the moment p, alpha2(p) is the sludge load rate of the secondary sedimentation tank at the moment p, alpha3(p) the concentration of suspended solids in the mixed solution of the secondary sedimentation tank at the moment p, alpha4(p) the retention time of the sludge in the secondary sedimentation tank at the moment p, alpha5(P) is a reflux ratio at the moment P, T is the transposition of the matrix, P is the moment of the training process of the fuzzy neural network, and P is the maximum iteration number;
radial base layer: the layer comprises k radial basis neurons, k is a positive integer between [5, 15], the input is blurred by a Gaussian function, and the output of each radial basis neuron is as follows:
wherein e is 2.72, betaj(p) is the output of the jth radial base neuron at time p, cij(p) is the central value of the ith input neuron and the jth radial base neuron at time p, σij(p) is the width value of the ith input neuron and the jth radial base neuron at time p, i ═ 1,2, …, 5, j ═ 1,2, …, k;
a normalization layer: this layer includes k normalization neurons, each of whose outputs are:
wherein etaj(p) is the output of the jth normalized neuron at time p, η (p) [. eta. ]1(p),η2(p),…,ηk(p)]TIs a normalized neuron output matrix;
an output layer: this layer includes 1 neuron, the output is:
g(p)＝w(p)η(p) (4)
where g (p) is the output value of the neural network, w (p) ═ w1(p),w2(p),…,wk(p)]Is a matrix of output weight parameters, wk(p) is the output weight of the kth normalized neuron at the time p;
(2) training a fuzzy neural network, wherein initial p is 1, and specifically:
calculating the fuzzy neural network error:
wherein g isd(p) is the desired output value of the fuzzy neural network;
updating parameters as follows:
wherein c isij(p +1) is the central value, σ, of the ith input neuron and the jth radial base neuron at time p +1ij(p +1) is the width value of the ith input neuron and the jth radial base neuron at time p +1, wj(p +1) is the output weight of the jth normalized neuron at the moment p + 1;
p is equal to P +1, if P is less than P, the steps of (r) - (c) are repeated, if P is equal to P, the cycle is ended;
(3) the secondary sedimentation tank is an important component part in the sewage treatment operation process, relevant parameters of a secondary sedimentation tank model are adjusted to obtain a transition model of a controlled object, and the method specifically comprises the following steps:
the double-exponential sludge settling rate function of the secondary sedimentation tank model can be described as follows:
wherein v (t) is the sludge sedimentation rate at time t, v0474m/d is the theoretical maximum sedimentation rate, rH＝0.000576m3the/g.SS is a sedimentation disturbance parameter, Xz(t) is the z-th layer suspension concentration at time t, z being 1,2, …, 10;
the method comprises the following steps of establishing a control model for recovering the normal working condition of sludge expansion according to different operating working conditions:
wherein
And
the change rates of the controlled variables dissolved oxygen concentration and nitrate nitrogen concentration at time t are shown, respectively, and u (t) ═ u1(t)，u2(t)]T，u1(t) and u2(t) control input variables aeration amount and internal reflux at time t, respectively, ftr(t) and btr(t) are respectively a 2 × 2 matrix and a 2-dimensional column vector, tr is the number of divided working conditions, and tr is 1,2, 3, and t is the time of the control process;
(4) the self-adaptive segmented sliding-mode control method for the dissolved oxygen concentration and the nitrate nitrogen concentration after the sludge bulking occurs is designed, and specifically comprises the following steps:
initializing three sliding mode controllers, each controller comprising a calculation S of a sliding mode surfacetr(t) solution of the control law utr(t) setting the initial value of t to 0;
calculating the deviation of the concentration of dissolved oxygen and the concentration of nitrate nitrogen of controlled variables:
e(t)＝x(t)-xd(t) (9)
wherein x (t) ═ x1(t)，x2(t)]T，x1(t) and x2(t) actual values, x, of the respective controlled variables dissolved oxygen concentration and nitrate nitrogen concentration at time td(t)＝[x1d(t)，x2d(t)]T，x1d(t) and x2d(t) respectively representing the expected values of the concentration of dissolved oxygen and the concentration of nitrate nitrogen of the controlled variable at the time t;
acquiring water quality data in a formula (1), namely the dissolved oxygen concentration of an aerobic tank, the sludge load rate of a secondary sedimentation tank, the suspended solid concentration of mixed liquid of the secondary sedimentation tank, the retention time and the reflux ratio of sludge of the secondary sedimentation tank, and calculating the sludge volume index at the moment according to formulas (2) to (4);
if the sludge volume index is between 0 and 150, executing the step three;
if the sludge volume index is more than 150 and less than or equal to 250, executing the step IV;
if the volume index of the sludge is more than 250, executing a fifth step;
the third, fourth and fifth steps are parallel;
calculating the sliding mode surface of the controller:
where e (τ) is the integrand on the deviation, d τ is the derivative of the integrating variable τ, k1The matrix is a 2 multiplied by 2 matrix, each element of the secondary diagonal is 0, and each element of the main diagonal is 0.035 and 0.25 respectively from top to bottom;
solving the control law of the controller:
wherein f is1(t) and b1F when (t) is tr-1tr(t) and btr(t) the corresponding variable(s),
and
respectively represents the change rate, lambda, of the expected values of the concentration of dissolved oxygen and the concentration of nitrate nitrogen of the controlled variable at the time t1(t) is an adaptive gain parameter at time t, which is related to the size of the sliding mode surface and can be expressed as 0.2 √ S1(t) |, always greater than 0, √ and | | | denote an evolution and an absolute value, respectively, sign () is a sign function whose expression is:
u to be solved1(t) assigning u (t) and then performing step (c);
fourthly, calculating the sliding mode surface of the controller:
wherein k is2Is a 2 x 2 matrix with 0 elements in the minor diagonal and-0.02, ζ from top to bottom in the major diagonal2＝[0.007，0.007]T；
Solving the control law of the controller:
wherein f is2(t) and b2F when (t) is tr-2tr(t) and btr(t) a variable corresponding to λ2(t) is an adaptive gain parameter at time t, which is related to the size of the sliding mode surface and can be expressed as 0.2 √ S2(t) |, always greater than 0;
u to be solved2(t) assigning u (t) and then performing step (c);
calculating the sliding mode surface of the controller:
wherein k is3Is a 2 × 2 matrix with 0 for each element of the minor diagonal and 0.025, 0.025 and zeta for each element of the major diagonal from top to bottom3＝[-0.013，-0.013]T；
Solving the control law of the controller:
wherein f is3(t) and b3F when (t) is tr-3tr(t) and btr(t) a variable corresponding to λ3(t) is an adaptive gain parameter at time t, which is related to the size of the sliding mode surface and can be expressed as 0.2 √ S3(t) |, always greater than 0;
to be solvedu3(t) assigning u (t) and then performing step (c);
sixthly, if t is t +1, if t is less than 300, the process goes to the step I, and if t is 300, the circulation is ended;
(5) and (5) performing tracking control on the dissolved oxygen concentration and the nitrate nitrogen concentration by using the solved u (t), wherein the final system output is an actual concentration value of the dissolved oxygen, an actual concentration value of the nitrate nitrogen and a predicted sludge volume index.
The invention is mainly characterized in that:
(1) the invention aims at the characteristic that the sewage treatment process is a complex industrial engineering with high occurrence rate of abnormal sludge expansion working conditions, the abnormal sludge expansion working conditions have the characteristics of difficult real-time detection and identification and the like, the fuzzy neural network is adopted to obtain the water quality parameters in the operation process and predict the sludge volume index to judge the working conditions of the sewage treatment operation, a reference signal is provided for the switching of the controller, and the invention has the characteristics of high efficiency, stability and the like.
(2) The invention adopts a sludge bulking suppression method based on self-adaptive segmented sliding mode control to regulate and control the dissolved oxygen concentration and nitrate nitrogen concentration in the sewage treatment process, and the method can judge the working condition according to the predicted sludge volume index, switch a proper sliding mode controller and coordinate the biochemical reaction process; the sewage treatment process can be transited to the normal working condition through a proper regulation and control method after the sludge expansion occurs.
Drawings
FIG. 1 is a general block diagram of the present invention
FIG. 2 is a graph showing the results of real-time sludge volume index prediction according to the present invention
FIG. 3 is a graph showing the results of controlling the concentration of dissolved oxygen in accordance with the present invention
FIG. 4 is a graph showing the control result of the nitrate nitrogen concentration according to the present invention
Detailed Description
The invention obtains a sludge bulking inhibition method based on self-adaptive segmented slip-form control, which predicts the sludge volume index of the sewage treatment process in real time by utilizing the dissolved oxygen concentration of an aerobic tank, the sludge load rate of a secondary sedimentation tank, the suspended solid concentration of mixed liquid of the secondary sedimentation tank, the retention time and the reflux ratio of sludge of the secondary sedimentation tank, constructs a multi-segment slip-form controller to adjust the dissolved oxygen concentration and the nitrate nitrogen concentration, and designs a proper switching mechanism to keep the sewage treatment process in a normal working condition and realize the inhibition of sludge bulking;
the invention adopts the following technical scheme and implementation steps:
1. a sludge bulking restraining method based on self-adaptive sectional sliding mode control,
the sludge volume index is predicted in real time, a proper sliding mode controller is designed, the inhibition of sludge expansion is completed, and the overall structure is shown in figure 1;
the method is characterized by comprising the following steps:
(1) a fuzzy neural network for real-time prediction of sludge volume index is designed, and the structure of the fuzzy neural network is divided into four layers:
the device comprises an input layer, a radial base layer, a normalization layer and an output layer; the method specifically comprises the following steps:
an input layer: this layer includes 5 input neurons:
α(p)＝[α1(p),α2(p),α3(p),α4(p),α5(p)]T (1)
where α (p) is the input vector of the fuzzy neural network at time p, α1(p) is the dissolved oxygen concentration of the aerobic tank at the moment p, alpha2(p) is the sludge load rate of the secondary sedimentation tank at the moment p, alpha3(p) the concentration of suspended solids in the mixed solution of the secondary sedimentation tank at the moment p, alpha4(p) the retention time of the sludge in the secondary sedimentation tank at the moment p, alpha5(P) is a reflux ratio at time P, T is the transposition of the matrix, P is the time of the training process of the fuzzy neural network, P is the maximum iteration number, and P is 700;
radial base layer: the layer includes k radial basis neurons, k 15, the input is blurred with a gaussian function, and the output of each radial basis neuron is:
wherein e is 2.72, betaj(p) j is the jth path at time pOutput to the base neuron, cij(p) is the central value of the ith input neuron and the jth radial base neuron at time p, σij(p) is the width value of the ith input neuron and the jth radial base neuron at time p, i ═ 1,2, …, 5, j ═ 1,2, …, k;
a normalization layer: this layer includes k normalization neurons, each of whose outputs are:
wherein etaj(p) is the output of the jth normalized neuron at time p, η (p) [. eta. ]1(p),η2(p),…,ηk(p)]TIs a normalized neuron output matrix;
an output layer: this layer includes 1 neuron, the output is:
g(p)＝w(p)η(p) (4)
where g (p) is the output value of the neural network, w (p) ═ w1(p),w2(p),…,wk(p)]Is a matrix of output weight parameters, wk(p) is the output weight of the kth normalized neuron at the time p;
(2) training a fuzzy neural network, wherein initial p is 1, and specifically:
calculating the fuzzy neural network error:
wherein g isd(p) is the desired output value of the fuzzy neural network;
updating parameters as follows:
wherein c isij(p +1) is the central value, σ, of the ith input neuron and the jth radial base neuron at time p +1ij(p +1) is p +Width values, w, of the ith input neuron and the jth radial base neuron at time 1j(p +1) is the output weight of the jth normalized neuron at the moment p + 1;
p is equal to P +1, if P is less than P, the steps of (r) - (c) are repeated, if P is equal to P, the cycle is ended;
(3) the secondary sedimentation tank is an important component part in the sewage treatment operation process, relevant parameters of a secondary sedimentation tank model are adjusted to obtain a transition model of a controlled object, and the method specifically comprises the following steps:
the double-exponential sludge settling rate function of the secondary sedimentation tank model can be described as follows:
wherein v (t) is the sludge sedimentation rate at time t, v0474m/d is the theoretical maximum sedimentation rate, rH＝0.000576m3the/g.SS is a sedimentation disturbance parameter, Xz(t) is the z-th layer suspension concentration at time t, z being 1,2, …, 10;
the method comprises the following steps of establishing a control model for recovering the normal working condition of sludge expansion according to different operating working conditions:
wherein
And
the change rates of the controlled variables dissolved oxygen concentration and nitrate nitrogen concentration at time t are shown, respectively, and u (t) ═ u1(t)，u2(t)]T，u1(t) and u2(t) control input variables aeration amount and internal reflux at time t, respectively, ftr(t) and btr(t) matrices of 2 x 2 and 2 dimensions, respectivelyThe column vector tr is the number of the divided working conditions, and is 1,2, 3, and t is the time of the control process;
(4) the self-adaptive segmented sliding-mode control method for the dissolved oxygen concentration and the nitrate nitrogen concentration after the sludge bulking occurs is designed, and specifically comprises the following steps:
initializing three sliding mode controllers, each controller comprising a calculation S of a sliding mode surfacetr(t) solution of the control law utr(t) setting the initial value of t to 0;
calculating the deviation of the concentration of dissolved oxygen and the concentration of nitrate nitrogen of controlled variables:
e(t)＝x(t)-xd(t) (9)
wherein x (t) ═ x1(t)，x2(t)]T，x1(t) and x2(t) actual values, x, of the respective controlled variables dissolved oxygen concentration and nitrate nitrogen concentration at time td(t)＝[x1d(t)，x2d(t)]T，x1d(t) and x2d(t) respectively representing the expected values of the concentration of dissolved oxygen and the concentration of nitrate nitrogen of the controlled variable at the time t;
acquiring water quality data in a formula (1), namely the dissolved oxygen concentration of an aerobic tank, the sludge load rate of a secondary sedimentation tank, the suspended solid concentration of mixed liquid of the secondary sedimentation tank, the retention time and the reflux ratio of sludge of the secondary sedimentation tank, and calculating the sludge volume index at the moment according to formulas (2) to (4);
if the sludge volume index is between 0 and 150, executing the step three;
if the sludge volume index is more than 150 and less than or equal to 250, executing the step IV;
if the volume index of the sludge is more than 250, executing a fifth step;
the third, fourth and fifth steps are parallel;
calculating the sliding mode surface of the controller:
where e (τ) is the integrand on the deviation, d τ is the derivative of the integrating variable τ, k1The matrix is a 2 multiplied by 2 matrix, each element of the secondary diagonal is 0, and each element of the main diagonal is 0.035 and 0.25 respectively from top to bottom;
solving the control law of the controller:
wherein f is1(t) and b1F when (t) is tr-1tr(t) and btr(t) the corresponding variable(s),
and
respectively represents the change rate, lambda, of the expected values of the concentration of dissolved oxygen and the concentration of nitrate nitrogen of the controlled variable at the time t1(t) is an adaptive gain parameter at time t, which is related to the size of the sliding mode surface and can be expressed as 0.2 √ S1(t) |, always greater than 0, √ and | | | denote an evolution and an absolute value, respectively, sign () is a sign function whose expression is:
u to be solved1(t) assigning u (t) and then performing step (c);
fourthly, calculating the sliding mode surface of the controller:
wherein k is2Is a 2 x 2 matrix with 0 elements in the minor diagonal and-0.02, ζ from top to bottom in the major diagonal2＝[0.007，0.007]T；
Solving the control law of the controller:
wherein f is2(t) and b2F when (t) is tr-2tr(t) and btr(t) a variable corresponding to λ2(t) is an adaptive gain parameter at time t, which is related to the size of the sliding mode surface and can be expressed as 0.2 √ S2(t) |, always greater than 0;
u to be solved2(t) assigning u (t) and then performing step (c);
calculating the sliding mode surface of the controller:
wherein k is3Is a 2 × 2 matrix with 0 for each element of the minor diagonal and 0.025, 0.025 and zeta for each element of the major diagonal from top to bottom3＝[-0.013，-0.013]T；
Solving the control law of the controller:
wherein f is3(t) and b3F when (t) is tr-3tr(t) and btr(t) a variable corresponding to λ3(t) is an adaptive gain parameter at time t, which is related to the size of the sliding mode surface and can be expressed as 0.2 √ S3(t) |, always greater than 0;
u to be solved3(t) assigning u (t) and then performing step (c);
sixthly, if t is t +1, if t is less than 300, the process goes to the step I, and if t is 300, the circulation is ended;
(5) and (5) performing tracking control on the dissolved oxygen concentration and the nitrate nitrogen concentration by using the solved u (t), wherein the final system output is an actual concentration value of the dissolved oxygen, an actual concentration value of the nitrate nitrogen and a predicted sludge volume index.
(5) Tracking and controlling the dissolved oxygen concentration and the nitrate nitrogen concentration by using the solved u (t), wherein the output of the final system is a predicted sludge volume index, an actual concentration value of dissolved oxygen and an actual concentration value of nitrate nitrogen; FIG. 2 shows the predicted value of the sludge volume index in the operation process of sewage treatment, wherein the X axis: time, in days, Y-axis: the unit of the predicted value of the sludge volume index is milliliter/gram, a black solid line is the upper limit of the sludge volume index under the normal working condition, and a black dotted line is the predicted value of the sludge volume index; fig. 3 shows the dissolved oxygen concentration values of the sewage treatment process, X-axis: time, in days, Y-axis: the dissolved oxygen concentration is in mg/l, the black solid line is the set value of the dissolved oxygen concentration, and the black dotted line is the actual value of the dissolved oxygen concentration; fig. 4 shows nitrate nitrogen concentration values in the wastewater treatment process, X-axis: time, in days, Y-axis: the unit of the nitrate nitrogen concentration is milligram/liter, a black solid line is a nitrate nitrogen concentration set value, and a black dotted line is a nitrate nitrogen concentration actual value; the experimental result proves the effectiveness of the method.
Claims (1)
1. A sludge bulking suppression method based on self-adaptive segmented slip-form control is characterized in that a proper slip-form controller is designed by predicting a sludge volume index in real time to complete suppression of sludge bulking;
the method is characterized by comprising the following steps:
(1) a fuzzy neural network for real-time prediction of sludge volume index is designed, and the structure of the fuzzy neural network is divided into four layers: the device comprises an input layer, a radial base layer, a normalization layer and an output layer; the method specifically comprises the following steps:
an input layer: this layer includes 5 input neurons:
α(p)＝[α1(p),α2(p),α3(p),α4(p),α5(p)]T (1)
where α (p) is the input vector of the fuzzy neural network at time p, α1(p) is the dissolved oxygen concentration of the aerobic tank at the moment p, alpha2(p) is the sludge load rate of the secondary sedimentation tank at the moment p, alpha3(p) the concentration of suspended solids in the mixed solution of the secondary sedimentation tank at the moment p, alpha4(p) is time pSludge retention time of secondary sedimentation tank, alpha5(P) is a reflux ratio at the moment P, T is the transposition of the matrix, P is the moment of the training process of the fuzzy neural network, and P is the maximum iteration number;
radial base layer: the layer comprises k radial basis neurons, k is a positive integer between [5, 15], the input is blurred by a Gaussian function, and the output of each radial basis neuron is as follows:
wherein e is 2.72, betaj(p) is the output of the jth radial base neuron at time p, cij(p) is the central value of the ith input neuron and the jth radial base neuron at time p, σij(p) is the width value of the ith input neuron and the jth radial base neuron at time p, i ═ 1,2, …, 5, j ═ 1,2, …, k;
a normalization layer: this layer includes k normalization neurons, each of whose outputs are:
wherein etaj(p) is the output of the jth normalized neuron at time p, η (p) [. eta. ]1(p),η2(p),…,ηk(p)]TIs a normalized neuron output matrix;
an output layer: this layer includes 1 neuron, the output is:
g(p)＝w(p)η(p) (4)
where g (p) is the output value of the neural network, w (p) ═ w1(p),w2(p),…,wk(p)]Is a matrix of output weight parameters, wk(p) is the output weight of the kth normalized neuron at the time p;
(2) training a fuzzy neural network, wherein initial p is 1, and specifically:
calculating the fuzzy neural network error:
wherein g isd(p) is the desired output value of the fuzzy neural network;
updating parameters as follows:
wherein c isij(p +1) is the central value, σ, of the ith input neuron and the jth radial base neuron at time p +1ij(p +1) is the width value of the ith input neuron and the jth radial base neuron at time p +1, wj(p +1) is the output weight of the jth normalized neuron at the moment p + 1;
p is equal to P +1, if P is less than P, the steps of (r) - (c) are repeated, if P is equal to P, the cycle is ended;
(3) the secondary sedimentation tank is an important component part in the sewage treatment operation process, relevant parameters of a secondary sedimentation tank model are adjusted to obtain a transition model of a controlled object, and the method specifically comprises the following steps:
the double-exponential sludge settling rate function of the secondary sedimentation tank model can be described as follows:
wherein v (t) is the sludge sedimentation rate at time t, v0474m/d is the theoretical maximum sedimentation rate, rH＝0.000576m3the/g.SS is a sedimentation disturbance parameter, Xz(t) is the z-th layer suspension concentration at time t, z being 1,2, …, 10;
the method comprises the following steps of establishing a control model for recovering the normal working condition of sludge expansion according to different operating working conditions:
wherein
And
the change rates of the controlled variables dissolved oxygen concentration and nitrate nitrogen concentration at time t are shown, respectively, and u (t) ═ u1(t)，u2(t)]T，u1(t) and u2(t) control input variables aeration amount and internal reflux at time t, respectively, ftr(t) and btr(t) is a 2 × 2 matrix and a 2-dimensional column vector respectively, tr is a label for dividing the working condition, and tr is 1,2, 3, and t is the time of the control process;
(4) the self-adaptive segmented sliding-mode control method for the dissolved oxygen concentration and the nitrate nitrogen concentration after the sludge bulking occurs is designed, and specifically comprises the following steps:
initializing three sliding mode controllers, each controller comprising a calculation S of a sliding mode surfacetr(t) solution of the control law utr(t) setting the initial value of t to 0;
calculating the deviation of the concentration of dissolved oxygen and the concentration of nitrate nitrogen of controlled variables:
e(t)＝x(t)-xd(t) (9)
wherein x (t) ═ x1(t)，x2(t)]T，x1(t) and x2(t) actual values, x, of the respective controlled variables dissolved oxygen concentration and nitrate nitrogen concentration at time td(t)＝[x1d(t)，x2d(t)]T，x1d(t) and x2d(t) respectively representing the expected values of the concentration of dissolved oxygen and the concentration of nitrate nitrogen of the controlled variable at the time t;
acquiring water quality data in a formula (1), namely the dissolved oxygen concentration of an aerobic tank, the sludge load rate of a secondary sedimentation tank, the suspended solid concentration of mixed liquid of the secondary sedimentation tank, the retention time and the reflux ratio of sludge of the secondary sedimentation tank, and calculating the sludge volume index at the moment according to formulas (2) to (4);
if the sludge volume index is between 0 and 150, executing the step three;
if the sludge volume index is more than 150 and less than or equal to 250, executing the step IV;
if the volume index of the sludge is more than 250, executing a fifth step;
the third, fourth and fifth steps are parallel;
calculating the sliding mode surface of the controller:
where e (τ) is the integrand on the deviation, d τ is the derivative of the integrating variable τ, k1The matrix is a 2 multiplied by 2 matrix, each element of the secondary diagonal is 0, and each element of the main diagonal is 0.035 and 0.25 respectively from top to bottom;
solving the control law of the controller:
wherein f is1(t) and b1F when (t) is tr-1tr(t) and btr(t) the corresponding variable(s),
and
respectively represents the change rate, lambda, of the expected values of the concentration of dissolved oxygen and the concentration of nitrate nitrogen of the controlled variable at the time t1(t) is an adaptive gain parameter at time t, which is related to the size of the sliding mode surface and can be expressed as 0.2 √ S1(t) |, always greater than 0, √ and | | | denote an evolution and an absolute value, respectively, sign () is a sign function whose expression is:
u to be solved1(t) assigning u (t) and then performing step (c);
fourthly, calculating the sliding mode surface of the controller:
wherein k is2Is a 2 x 2 matrix with 0 elements in the minor diagonal and-0.02, ζ from top to bottom in the major diagonal2＝[0.007，0.007]T；
Solving the control law of the controller:
wherein f is2(t) and b2F when (t) is tr-2tr(t) and btr(t) a variable corresponding to λ2(t) is an adaptive gain parameter at time t, which is related to the size of the sliding mode surface and can be expressed as 0.2 √ S2(t) |, always greater than 0;
u to be solved2(t) assigning u (t) and then performing step (c);
calculating the sliding mode surface of the controller:
wherein k is3Is a 2 × 2 matrix with 0 for each element of the minor diagonal and 0.025, 0.025 and zeta for each element of the major diagonal from top to bottom3＝[-0.013，-0.013]T；
Solving the control law of the controller:
wherein f is3(t) and b3F when (t) is tr-3tr(t) and btr(t) a variable corresponding to λ3(t) is an adaptive gain parameter at time t, which is related to the size of the sliding mode surface and can be expressed as 0.2 √ S3(t) |, always greater than 0;
u to be solved3(t) assigning u (t) and then performing step (c);
sixthly, if t is t +1, if t is less than 300, the process goes to the step I, and if t is 300, the circulation is ended;
(5) and (5) performing tracking control on the dissolved oxygen concentration and the nitrate nitrogen concentration by using the solved u (t), wherein the final system output is an actual concentration value of the dissolved oxygen, an actual concentration value of the nitrate nitrogen and a predicted sludge volume index.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011092548.5A CN112363391B (en) | 2020-10-14 | 2020-10-14 | Sludge bulking inhibition method based on self-adaptive segmented sliding mode control |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011092548.5A CN112363391B (en) | 2020-10-14 | 2020-10-14 | Sludge bulking inhibition method based on self-adaptive segmented sliding mode control |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112363391A true CN112363391A (en) | 2021-02-12 |
| CN112363391B CN112363391B (en) | 2022-03-11 |
Family
ID=74508131
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011092548.5A Active CN112363391B (en) | 2020-10-14 | 2020-10-14 | Sludge bulking inhibition method based on self-adaptive segmented sliding mode control |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112363391B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113031445A (en) * | 2021-03-12 | 2021-06-25 | 北京工业大学 | Mechanism model-based robust multivariable control method for sewage denitrification process |
| CN113391554A (en) * | 2021-06-16 | 2021-09-14 | 周节 | Electroplating method based on artificial intelligence |
| CN114859711A (en) * | 2022-04-15 | 2022-08-05 | 北京工业大学 | Urban sewage treatment aeration process control method based on fuzzy second-order sliding mode |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102156406A (en) * | 2011-02-28 | 2011-08-17 | 哈尔滨工业大学 | Expert system and method for preventing and controlling sludge bulking under the diagnosis based on operation condition of sewage disposal plant |
| CN103880183A (en) * | 2014-03-27 | 2014-06-25 | 北京工业大学 | Real-time control system and method of A<2>/O-biological contact oxidation device for shortcut nitrification coupled denitrifying phosphorus removal |
| CN108898215A (en) * | 2018-07-18 | 2018-11-27 | 北京工业大学 | A kind of sludge bulking INTELLIGENT IDENTIFICATION method based on two type fuzzy neural networks |
-
2020
- 2020-10-14 CN CN202011092548.5A patent/CN112363391B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102156406A (en) * | 2011-02-28 | 2011-08-17 | 哈尔滨工业大学 | Expert system and method for preventing and controlling sludge bulking under the diagnosis based on operation condition of sewage disposal plant |
| CN103880183A (en) * | 2014-03-27 | 2014-06-25 | 北京工业大学 | Real-time control system and method of A<2>/O-biological contact oxidation device for shortcut nitrification coupled denitrifying phosphorus removal |
| CN108898215A (en) * | 2018-07-18 | 2018-11-27 | 北京工业大学 | A kind of sludge bulking INTELLIGENT IDENTIFICATION method based on two type fuzzy neural networks |
Non-Patent Citations (1)
| Title |
|---|
| 李素康 等: ""污水处理曝气过程中溶解氧浓度智能控制系统"", 《中国新技术新产品》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113031445A (en) * | 2021-03-12 | 2021-06-25 | 北京工业大学 | Mechanism model-based robust multivariable control method for sewage denitrification process |
| CN113031445B (en) * | 2021-03-12 | 2022-09-09 | 北京工业大学 | Mechanism model-based robust multivariable control method for sewage denitrification process |
| CN113391554A (en) * | 2021-06-16 | 2021-09-14 | 周节 | Electroplating method based on artificial intelligence |
| CN113391554B (en) * | 2021-06-16 | 2022-06-17 | 江苏东南环保科技有限公司 | Electroplating method based on artificial intelligence |
| CN114859711A (en) * | 2022-04-15 | 2022-08-05 | 北京工业大学 | Urban sewage treatment aeration process control method based on fuzzy second-order sliding mode |
| CN114859711B (en) * | 2022-04-15 | 2024-07-16 | 北京工业大学 | Urban sewage treatment aeration process control method based on fuzzy second-order sliding mode |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112363391B (en) | 2022-03-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112363391B (en) | Sludge bulking inhibition method based on self-adaptive segmented sliding mode control | |
| CN106295800B (en) | A kind of water outlet total nitrogen TN intelligent detecting method based on recurrence Self organizing RBF Neural Network | |
| Gontarski et al. | Simulation of an industrial wastewater treatment plant using artificial neural networks | |
| CN108445757B (en) | Sewage treatment process optimization control method based on dynamic multi-target particle swarm algorithm | |
| CN109669352A (en) | Oily waste water treatment procedure optimization control method based on adaptive multi-objective particle swarm | |
| CN115356930B (en) | Multi-objective optimization control system and method in sewage treatment process | |
| CN106406094A (en) | Interval type-2 fuzzy neural network based dissolved oxygen concentration tracking method in sewage treatment | |
| Vlad et al. | Indirect control of substrate concentration for a wastewater treatment process by dissolved oxygen tracking | |
| CN109782588A (en) | The tight format non-model control method of the different factor of MIMO | |
| Yasmin et al. | Performance comparison of SVM and ANN for aerobic granular sludge | |
| CN115097886A (en) | Method, system, equipment and medium for controlling concentration of dissolved oxygen in sewage treatment | |
| Fu et al. | Dissolved oxygen control system based on the TS fuzzy neural network | |
| Fan et al. | Optimization control of SBR wastewater treatment process based on pattern recognition | |
| CN116679026B (en) | Self-adaptive unbiased finite impulse response filtering sewage dissolved oxygen concentration estimation method | |
| CN111399558B (en) | Knowledge selection-based multi-objective optimization control method for sewage treatment process | |
| Qiao et al. | Recurrent neural network-based control for wastewater treatment process | |
| CN110412878A (en) | The estimation of PID-RBF stable state and LSSVM hard measurement sewage BOD control method | |
| Wang et al. | Kalman Filter based Newton Extremum Seeking Control for Maximum Gases Production Rates of Anaerobic Digestion Process | |
| Aglodiya | Application of artificial neural network (ANN) in chemical engineering: a review | |
| Husin et al. | Improving total nitrogen removal using a neural network ammonia-based aeration control in activated sludge process | |
| Diehl et al. | Fundamental nonlinearities of the reactor-settler interaction in the activated sludge process | |
| Barbusiński et al. | Simulation of the influence of wastewater quality indicators and operating parameters of a bioreactor on the variability of nitrogen in outflow and bulking of sludge: Data mining approach | |
| Juuso et al. | Intelligent analysers and dynamic simulation in a biological water treatment process | |
| Normandin et al. | Optimizing control of a continuous stirred tank fermenter using a neural network | |
| Zhang et al. | Multi-Objective Optimal Control of Wastewater Treatment Processes Constrainted by Energy-Saving and Effluent Qualities |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |