Disclosure of Invention
The invention provides a deep-sea sewage optimized discharge method based on tidal current change for solving the technical problems in the known technology.
The technical scheme adopted by the invention for solving the technical problems in the prior art is as follows: a sewage deep sea optimized discharge method based on tidal current change is characterized in that tidal current monitoring equipment is installed on a sewage discharge sea diffuser or at a position adjacent to the diffuser; establishing a response relation between the flow speed in the tidal current change interval of the engineering sea area and the optimal sewage discharge amount corresponding to the flow speed in the tidal current change interval of the engineering sea area in a pollution discharge control system:
(Qmin,Q1,…,Qi,…,Qn,Qmax)=F(umin,u1,…,ui,…,un,umax)
in the formula (u)min,u1,…,ui,…,un,umax) Acquiring the flow velocity in the tidal current change interval of the engineering sea area by collecting historical actual measurement data of the engineering and the tidal current of the adjacent sea area; (Q)min,Q1,…,Qi,…,Qn,Qmax) The optimal sewage discharge amount corresponding to the tidal current flow velocity in the interval is obtained; f is the response relation between the two;
the sewage discharge control system receives the actually measured flow velocity u ' data of the tide in the area near the sewage discharge port, which is obtained by the real-time monitoring of the tide monitoring equipment, inputs the actually measured flow velocity u ' data into the response relation, obtains the real-time optimal sewage discharge amount aiming at the actually measured flow velocity u ' of the tide, and sends a sewage discharge amount instruction to the sewage pump according to the real-time optimal sewage discharge amount;
the optimal sewage discharge amount in the response relation is calculated and obtained by adopting a simulation inverse algorithm, and the method comprises the following specific steps:
1) establishing a simulation back calculation method model
1.1) adopting a sewage outlet pollutant diffusion mathematical model;
1.2) establishing a Multi-objective function
MAX.DD=C0/C1 (1)
MIN.EC=0.0414×Q+13.828 (2)
MAX.OE=OE (3)
In the formula (1), DD is the dilution of pollutants; c0The concentration value of the pollutants in the jet flow water body is obtained; c1The concentration value of the pollutant at a certain point in the environmental water body is obtained; EC is energy consumption value; q is the sewage discharge amount of the pump, m is m multiplied by v multiplied by S, m is the number of the nozzles, v is the discharge rate, and S is the sectional area of the nozzles; OE is the sewage discharge efficiency;
1.3) Using a genetic Algorithm model
1.3.1) calculation operator by evolution
1.3.2) using cross-computing operators
1.3.3) calculation operator by mutation
2) Initial setting
2.1) taking the sewage discharge as a control variable of an optimization process, and adopting a real number coding form;
2.2) setting a value range of the control variable, and taking the value range as a constraint condition of the genetic algorithm population value;
3) performing the simulation inverse calculation of the optimal sewage discharge amount according to the following steps aiming at each tidal flow velocity in the n +2 tidal flow velocities, and obtaining n +2 optimal sewage discharge amounts aiming at the n +2 tidal flow velocities
3.1) randomly selecting r numerical values in the control variable value range according to each tidal current flow rate to obtain r chromosomes, and forming a population (r) of the possible solution of the optimal sewage discharge amount under the tidal current flow rate condition;
3.2) taking each chromosome as a numerical value example, adopting a pollutant diffusion mathematical model to calculate r times, and simultaneously outputting and storing r pollutant concentration results of observation points in the water area;
3.3) carrying out chromosome evaluation by adopting the multi-target function established in the step 1.2) with the target of optimal sewage dilution effect, lowest energy consumption and optimal sewage discharge efficiency;
3.4) when the maximum generation number is met, jumping out of the program and outputting the optimal chromosome, namely the optimal sewage discharge amount; if the maximum generation number target is not met, adopting the genetic algorithm model in the step 1.3) to perform genetic calculation;
3.5) updating the chromosomes selected in the step 3.1) through genetic calculation, obtaining a brand new population, and repeating the calculation according to the steps (3.2) to (3.4) until the best chromosomes, namely the optimal sewage discharge amount, is output.
The method for establishing the response relation comprises the following steps:
1) establishing a data-driven BP algorithm model of the artificial neural network, wherein a hyperbolic Sigmoid function is adopted as a nonlinear conversion function in the BP algorithm model of the artificial neural network, and the formula (9) is as follows:
f(x)=1/1+e-x+θ (9)
in the formula, theta represents a threshold value;
in the artificial neural network BP algorithm model, Root Mean Square Error (RMSE) is adopted to evaluate the prediction capability of the network, and the formula (10) shows:
in the formula, a is the number of data; y isiIs an actual measurement value; y'iIs a predicted value;
the normalization processing of all data in the artificial neural network BP algorithm model adopts an equation (11):
wherein:
represents a normalized value; y is
iRepresents the value before normalization; y is
maxRepresents the maximum of all data; y is
minRepresents the minimum of all data;
2) initializing network weight, setting a network architecture by combining n +2 tidal current flow rates and optimal sewage discharge amount corresponding to the tidal current flow rates, and setting basic parameters of an artificial neural network BP algorithm model;
3) taking n +2 tidal current flow rate data as the input of an artificial neural network BP algorithm model, correspondingly taking n +2 optimal sewage discharge amount as the output of the artificial neural network BP algorithm model, substituting the optimal sewage discharge amount into the artificial neural network model for training and learning, and calculating the root mean square error;
4) when the error requirement of Root Mean Square Error (RMSE) is met, a jump-out program establishes a response relation; if the Root Mean Square Error (RMSE) error requirement is not met, performing subsequent calculation;
5) and (4) based on the root mean square error, after the network weight is reversely adjusted, repeating the calculation according to the steps 2) to 4) until a response relation is established.
The invention has the advantages and positive effects that: by taking the optimal sewage dilution effect, the lowest energy consumption and the optimal sewage discharge efficiency as targets, the response relation between the flow speed in the tidal current change interval of the engineering sea area and the optimal sewage discharge amount corresponding to the flow speed is established, the response relation is applied, the actual measurement tidal current flow speed is taken as the basis, the deep sea sewage discharge amount is optimized and adjusted in real time, the sewage is highly diluted in a space-time range as small as possible, the initial dilution effect is improved, the sewage is finally discharged up to the standard, the management level of the deep sea sewage discharge can be improved, the ecological quality level of the sea area environment is maintained, and the method has certain protection effect and practical significance on the ecological quality of the engineering and the adjacent sea area environment. The invention also has the characteristics of low cost, good economy, no need of a large amount of engineering construction cost, high fusion degree of the existing engineering and remarkable environmental and ecological benefits.
Detailed Description
In order to further understand the contents, features and effects of the present invention, the following embodiments are illustrated and described in detail with reference to the accompanying drawings:
referring to fig. 1, in a sewage deep sea drainage system, a sewage pump station 1 is provided with a sewage control system 2 and a sewage pump 3, the sewage pump 3 is connected with a sewage pipeline 4, a drainage section of the sewage pipeline 4 is located below a sea level 7m, a plurality of sewage drainage sea diffusers 5 are arranged on the drainage section of the sewage pipeline 4, and a nozzle is arranged on each sewage drainage sea diffuser 5 and is arranged horizontally.
The discharge system adopts a deep-sea sewage optimized discharge method based on tidal current change, based on regional tidal current monitoring at the deep-sea sewage discharge opening, and based on a sewage discharge control system in a land sewage pump station as a main tool, the sewage discharge control system 2 calculates the optimal sewage discharge amount in real time, sends a working instruction to the sewage pump 3 and controls the working efficiency of the sewage pump, and the specific measures are as follows: installing a tidal current monitoring device 6 on or adjacent to the sewage drainage diffuser 5; establishing a response relation between the flow speed in the tidal current change interval of the engineering sea area and the optimal sewage discharge amount corresponding to the flow speed in the sewage discharge control system 2:
(Qmin,Q1,…,Qi,…,Qn,Qmax)=F(umin,u1,…,ui,…,un,umax)
in the formula (u)min,u1,…,ui,…,un,umax) Acquiring the flow velocity in the tidal current change interval of the engineering sea area by collecting historical actual measurement data of the engineering and the tidal current of the adjacent sea area; (Q)min,Q1,…,Qi,…,Qn,Qmax) The optimal sewage discharge amount corresponding to the tidal current flow velocity in the interval is obtained; f is the response relationship between the two.
The sewage discharge control system 2 receives the actually measured tidal current flow velocity u ' data of the area near the sewage discharge port, which is obtained by the real-time monitoring of the tidal current monitoring equipment 6, inputs the actually measured tidal current flow velocity u ' data into the response relation, obtains the real-time optimal sewage discharge amount aiming at the actually measured tidal current flow velocity u ', and sends a sewage discharge amount instruction to the sewage pump 3 according to the real-time optimal sewage discharge amount.
Before the tidal current monitoring equipment 6 is installed, historical actual measurement data of tidal current of engineering and adjacent sea areas need to be collected, basic characteristics such as tidal current types and flow field distribution of the engineering sea areas are analyzed and grasped by combining a tidal current dynamic numerical simulation technology of the engineering sea areas, the maximum value and the minimum value of tidal current flow rate are extracted, and a tidal current change interval and relevant position information of the engineering sea areas are determined. According to basic characteristics such as the type and the flow field distribution of the tidal current in the engineering sea area and position information related to the maximum value and the minimum value of the tidal current flow rate, and according to basic principles such as strong region representativeness, simple construction, stability and reliability, a proper point position is selected on a sewage drainage diffuser or an adjacent position to be used as a tidal current monitoring point position, and the tidal current monitoring equipment 6 is installed on the tidal current monitoring point position.
The optimal sewage discharge amount in the response relation is calculated by adopting a simulation inverse algorithm, and the method comprises the following specific steps:
1) establishing a simulation back calculation method model
1.1) the mathematical model of the diffusion of pollutants at sewage outlets is adopted, please refer to the following two documents, namely, first) the numerical simulation of the influence of the Jet angle on the Jet characteristics in the environment of Flowing water [ J ]. university of continental Engineering, 2007,47(4): 583-.
1.2) establishing a Multi-objective function
MAX.DD=C0/C1 (1)
MIN.EC=0.0414×Q+13.828 (2)
MAX.OE=OE (3)
In the formula (1), DD is the dilution of pollutants; c0The concentration value of the pollutants in the jet flow water body is obtained; c1The concentration value of the pollutant at a certain point in the environmental water body is obtained; EC is energy consumption value; q is the sewage discharge amount of the pump, m is m multiplied by v multiplied by S, m is the number of the nozzles, v is the discharge rate, and S is the sectional area of the nozzles; OE is sewage discharge efficiency, and for practical engineers, the higher the efficiency, the better; the formula (2) is obtained by adopting least square fitting according to the flow and the energy consumption of the Germany KSB company K400-500 type horizontal sewage pump under different operation modes.
1.3) Using a genetic Algorithm model
The genetic algorithm applies the prior art, comprises three calculation parts, namely evolution, intersection and variation, and specifically comprises a genetic algorithm model:
1.3.1) calculation operator by evolution
1.3.2) using cross-computing operators
1.3.3) calculation operator by mutation
The genetic algorithm comprises the following steps: evolutionary calculations (roulette selection), cross calculations (as shown in equations (4) and (5)), and variant calculations (as shown in equations (6) to (8)).
Os1=ω×Pa1+(1-ω)×Pa2 (4)
Os2=ω×Pa2+(1-ω)×Pa1 (5)
In the formulas (4) and (5), omega is a random parameter between-0.25 and 1.25; os and Pa represent offspring and father, respectively; 1 and 2 are labels for children and parents;
and (3) calculating variation:
X=X'+Δ(t,y) (6)
in the formulas (6) and (7), X is a mutated gene, X' is an initial gene, T is the current generation number, T is the maximum generation number, r is a random number between 0 and 1, and b is a system parameter. The parameter y takes the value according to the formula (8), LD is the minimum value of the gene, UD is the maximum value of the gene, and i is a random number which is 0 or 1 respectively.
2) Initial setting
2.1) taking the sewage discharge amount as a control variable of an optimization process, and adopting a real number coding form.
And 2.2) setting the value range of the control variable, and taking the value range as a constraint condition of the population value of the genetic algorithm.
3) For each of the n +2 tidal flow velocities, performing analog back calculation of the optimal sewage discharge amount according to the following steps to obtain n +2 optimal sewage discharge amounts for the n +2 tidal flow velocities, wherein the specific steps are as follows:
3.1) randomly selecting r numerical values in the control variable value range according to each tidal current flow velocity to obtain r chromosomes, and forming a population (r) of the possible solution of the optimal sewage discharge amount under the tidal current flow velocity condition.
And 3.2) taking each chromosome as a numerical example, adopting a pollutant diffusion mathematical model to perform r times of calculation, and simultaneously outputting and storing r pollutant concentration results of observation points in the water area.
3.3) carrying out chromosome evaluation by adopting the multi-objective function established in the step 1.2) with the objectives of optimal sewage dilution effect, lowest energy consumption and optimal sewage discharge efficiency.
3.4) when the maximum generation number is met, jumping out of the program and outputting the optimal chromosome, namely the optimal sewage discharge amount; if the maximum generation number target is not met, adopting the genetic algorithm model in the step 1.3) to perform genetic calculation.
3.5) updating the chromosomes selected in the step 3.1) through genetic calculation, obtaining a brand new population, and repeating the calculation according to the steps (3.2) to (3.4) until the best chromosomes, namely the optimal sewage discharge amount, is output.
The simulation inverse algorithm embeds a sewage discharge port pollutant diffusion mathematical model (also called a jet flow model) and an operation energy consumption formula into a genetic algorithm model, takes the optimal sewage dilution effect, the lowest energy consumption and the optimal discharge efficiency as targets, carries out a coupling method model of optimal sewage discharge amount simulation inverse calculation, adopts the sewage discharge port pollutant diffusion mathematical model to calculate each chromosome and outputs pollutant concentration data of an observation point in a calculated water area; calculating and analyzing the dilution, sewage conveying energy consumption and discharge efficiency of each chromosome by adopting a multi-objective function; taking the maximum generation number set by the model as a judgment condition, jumping out of a calculation program when the condition is met, and entering genetic calculation when the condition is not met; and (4) adjusting chromosomes by genetic calculation operators of evolution, intersection and variation to form a brand new population, and substituting the brand new population into a pollutant diffusion mathematical model of the sewage outlet to perform calculation of the next generation.
The method for establishing the response relation comprises the following steps:
1) establishing a data-driven BP algorithm model of the artificial neural network, wherein a hyperbolic Sigmoid function is adopted as a nonlinear conversion function in the BP algorithm model of the artificial neural network, and the formula (9) is as follows:
f(x)=1/1+e-x+θ (9)
in the formula, theta represents a threshold value;
in the artificial neural network BP algorithm model, Root Mean Square Error (RMSE) is adopted to evaluate the prediction capability of the network, and the formula (10) shows:
in the formula, a is the number of data; y isiIs an actual measurement value; y isi' is a predicted value.
The normalization processing of all data in the artificial neural network BP algorithm model adopts an equation (11):
wherein:
represents a normalized value; y is
iBefore showing normalizationA value; y is
maxRepresents the maximum of all data; y is
minRepresents the minimum of all data.
2) Initializing network weight, setting a network architecture by combining n +2 tidal current flow rates and the optimal sewage discharge amount corresponding to the tidal current flow rates, and setting basic parameters of an artificial neural network BP algorithm model. The determination of the basic parameters can be found in the literature: lemingchang, Zhang Guangyu, Qiqi, and the like, a data-driven and genetic-computation-based sea-area combined unit water quality model multi-parameter step-by-step coupling optimization inversion method researches [ J ] practice and knowledge of mathematics, 2015,45(12): 167-175.
3) And taking n +2 tidal current flow rate data as the input of the BP algorithm model of the artificial neural network, correspondingly taking n +2 optimal sewage discharge amount as the output of the BP algorithm model of the artificial neural network, substituting the optimal sewage discharge amount into the artificial neural network model for training and learning, and calculating the root mean square error.
4) When the error requirement of Root Mean Square Error (RMSE) is met, a jump-out program establishes a response relation; if the Root Mean Square Error (RMSE) error requirement is not met, then the subsequent calculation is entered.
5) And (4) based on the root mean square error, after the network weight is reversely adjusted, repeating the calculation according to the steps 2) to 4) until a response relation is established.
The response relation establishing method has the advantages that the data-driven model-artificial neural network BP algorithm is adopted, and the response relation between the system variables can be established only by taking the system state variables and the control variables as model input and output and analyzing the characteristics of the system data on the basis of little knowledge of the system physical knowledge.
Although the preferred embodiments of the present invention have been described above with reference to the accompanying drawings, the present invention is not limited to the above-described embodiments, which are merely illustrative and not restrictive, and those skilled in the art can make many modifications without departing from the spirit and scope of the present invention as defined in the appended claims.