Search

CN-121997807-A - Large-scale sea water pollution traceability monitoring site optimization layout method coupling efficient optimization and fine simulation

CN121997807ACN 121997807 ACN121997807 ACN 121997807ACN-121997807-A

Abstract

The invention provides an optimized layout method for a large-scale sea area water pollution tracing monitoring station, which is coupled with efficient optimization and fine simulation, and belongs to the technical field of water pollution source identification. The method comprises the steps of constructing a three-dimensional hydrodynamic force-water quality coupling model by adopting numerical simulation software, determining key decision variables of the three-dimensional hydrodynamic force-water quality coupling model through Sobol sensitivity analysis, initializing a radial basis function proxy model, carrying out iterative solution on the radial basis function proxy model to obtain pollutant time sequence concentration, and obtaining a monitoring station layout scheme by taking maximum tracing precision and minimum layout cost as targets based on the pollutant time sequence concentration. The method simplifies the three-dimensional hydrodynamic force-water quality coupling model through the radial basis function proxy model, has lower calculation consumption, and can effectively optimize the high-dimensional problem. And calculating monitoring precision based on the pollutant time sequence concentration predicted by the model and the actual pollutant concentration so as to optimize the layout scheme of the monitoring station.

Inventors

  • WEI GUOZHEN
  • Guan Xinxiong
  • Fan xinfei
  • XU YUANLU

Assignees

  • 大连海事大学

Dates

Publication Date
20260508
Application Date
20251230

Claims (10)

  1. 1. A large-scale sea area water pollution tracing monitoring site optimization layout method for coupling efficient optimization and fine simulation is characterized by comprising the following steps: Adopting numerical simulation software to construct a three-dimensional hydrodynamic force-water quality coupling model; determining key decision variables of the three-dimensional hydrodynamic-water quality coupling model through Sobol sensitivity analysis; Initializing a radial basis function proxy model; carrying out iterative solution on the radial basis function proxy model to obtain the pollutant time sequence concentration; and obtaining a monitoring station layout scheme based on the pollutant time sequence concentration and taking the maximized tracing precision and the minimized layout cost as targets.
  2. 2. The method of claim 1, wherein the three-dimensional hydrodynamic-water quality coupling model comprises: Wherein Q represents a source or sink amount per unit area; 、 The conversion coefficient between the orthogonal curve coordinate system and the Cartesian coordinate system is usually used for correcting the influence caused by the change of the coordinates when the conversion coefficient is usually generated in the curve coordinate system; Represents the horizontal coordinate in the curve coordinate system, U is V is the velocity component of η; Indicating the water level change of the water surface relative to the reference water level, i.e. the free water surface elevation, Representing the resting water depth relative to a reference water level; momentum equation: horizontal level In the direction of: horizontal level In the direction of: in the formula, Is the hydrostatic pressure gradient in the direction, To at the same time A hydrostatic pressure gradient in the direction; Representation of Imbalance of reynolds stress in the direction of imbalance of reynolds stress in the direction, Representation of Imbalance in Reynolds stress in the direction; Is the vortex-induced viscosity coefficient in the vertical direction; is the density of the water body; And The momentum change caused by external source is represented by the momentum influence of drainage, water withdrawal, wind stress and bottom shear stress, f is the Coriolis force coefficient for describing the deflection force caused by the rotation of earth, u, v, Respectively expressed in an orthogonal curve coordinate system 、 、 Speed change values in three directions, t represents time; is defined in motion The vertical velocity of the space is as follows The coordinate system is obtained by the following continuous equation: Wherein, the In order for an inflow item to be entered, Is an outflow item.
  3. 3. The method of claim 1, wherein said determining key decision variables of said three-dimensional hydrodynamic-water quality coupling model by Sobol sensitivity analysis comprises: The traceability decision variables comprise pollution source positions and time sequence emission concentrations of various pollutants; And calculating sensitivity indexes of all the traceable decision variables, wherein the traceable decision variables with sensitivity indexes larger than a preset threshold value are used as key decision variables.
  4. 4. The method of claim 1, wherein initializing the radial basis function proxy model comprises: generating a set of initial sample points in a parameter space based on a Latin hypercube design; Calculating an objective function value of an initial sample point through a three-dimensional hydrodynamic force-water quality coupling model; And constructing an initialized radial basis function proxy model based on the objective function value of the initial sample point.
  5. 5. The method of claim 4, wherein the objective function comprises: Wherein, the 、 Respectively a weight vector and a polynomial coefficient; is a linear polynomial at d variables.
  6. 6. The method of claim 4, wherein iteratively solving the radial basis function proxy model achieves water pollution tracing, comprising: generating candidate points on the radial basis function proxy model through an optimization algorithm; Predicting an objective function value of the candidate point using the radial basis function proxy model; calculating the minimum Euclidean distance between the candidate point and all the evaluated points; carrying out weighted summation on the minimum Euclidean distance and the objective function value to obtain a candidate point comprehensive score; Iteration is carried out on the radial basis function proxy model by taking candidate points with the comprehensive scores of the candidate points being larger than a preset threshold value as evaluation points of the next round; and outputting the optimal solution until the maximum evaluation times or convergence conditions are reached.
  7. 7. The method of claim 1, wherein obtaining a monitoring site deployment scenario based on the contaminant time series concentration targeting maximizing traceability accuracy, minimizing deployment cost, comprises: generating a plurality of monitoring point layout schemes based on existing monitoring points in a target sea area; constructing a total evaluation index based on the tracing precision and the layout cost; and calculating the total evaluation index of each scheme, wherein the layout scheme with the maximum total evaluation index is the final monitoring station layout scheme.
  8. 8. The method of claim 1, wherein the monitoring site arrangement satisfies physical constraints, comprising: The water depth of the area where the monitoring point is located is greater than the minimum operation depth; The area where the monitoring point is located is not in the main channel; the area where the monitoring point is located houses the power and communication infrastructure.
  9. 9. The method of claim 1, wherein the layout cost: The product of the number of monitoring sites and the cost of a single detection site; The tracing precision is as follows: Wherein, the For the sum of absolute errors between the time-series concentration values of the contaminants output by the model and the actual observed values, Is the standard deviation between actual observations.
  10. 10. The method of claim 1, wherein obtaining a monitoring site deployment scenario based on the contaminant time series concentration targeting maximizing traceability accuracy, minimizing deployment cost, comprises: calculating a tracing precision index: calculating an economic cost index: The decision variable M can be expressed as: M=Mj= (1 j ) As deployable within an area Is the sum of (3); normalizing the tracing precision index and the economic cost index: Wherein, the For the minimum value of the tracing precision index, For the maximum value of the tracing precision index, At the maximum value of the economic cost, Is the minimum value of economic cost; And carrying out weighted fusion on the normalized result to serve as a total evaluation index: Wherein, the , Is a weight coefficient.

Description

Large-scale sea water pollution traceability monitoring site optimization layout method coupling efficient optimization and fine simulation Technical Field The invention relates to the technical field of water pollution source identification, in particular to an optimized layout method for a large-scale sea area water pollution tracing monitoring station for coupling efficient optimization and fine simulation. Background In recent years, with the continued development of economic and human living standards, water pollution events frequently occur, including transportation accidents, sewage pipe bursts, discharge of illegal sewage treatment plants, terrorist attacks, and extreme natural disasters (such as earthquakes), etc. These events cause serious damage to the sustainable development of the environment, economy and society. Therefore, how to quickly and accurately identify the water pollution source based on the prior art so as to reduce the negative influence thereof becomes a problem to be solved urgently. Pollution source identification is generally regarded as an inverse problem of discomfort, involves complex physical, chemical, biological reactions, and is closely related to hydrodynamic conditions. The problem has been studied in various water systems such as channel irrigation systems, groundwater systems, water supply networks, and water systems. To solve this problem, various methods have been proposed, including classical regularization methods, optimization algorithms, and bayesian inference methods. In early researches, tikhonov regularization method was widely used as one of classical regularization methods, but because it relies on a simple physical model, it is difficult to deal with the problem of pollution source identification in complex water systems. With the development of computer technology, optimization algorithms (such as genetic algorithm, heuristic harmonious search algorithm, particle swarm optimization) are used for pollution source identification in combination with water quality simulation models, and errors between simulation values and observation values are reduced mainly by adjusting decision variables. However, these methods are computationally expensive and require a large number of simulations. Bayesian inference methods, while providing a range of distribution of pollution sources, also suffer from low computational efficiency. To reduce the computational cost, proxy models (such as polynomial regression, kriging) were introduced, but in high-dimensional nonlinearity problems, local optima were easily trapped and dynamic exploration was inadequate. In addition, although the parallelization technology can shorten the calculation time, the existing research has rarely applied the parallelization technology to the problem of pollution source identification by combining the parallelization technology with a proxy-optimization method. Although the large-scale sea water pollution source tracing technology has advanced, the identification precision of the large-scale sea water pollution source tracing technology is seriously dependent on the number and the spatial distribution of monitoring stations, and the precise tracing needs to optimize the monitoring layout. Traditional monitoring layout is based on expert experience, historical data statistics or simple space uniformity principle, and the 'traceability' and 'layout' are not deeply coupled, so that subjectivity is high, blindness is high, and accurate traceability requirements are difficult to meet. Therefore, an innovative technical scheme is needed, a refined water environment-hydrodynamic model, a high-efficiency optimization algorithm and monitoring station layout optimization are tightly combined, and a closed-loop and self-adaptive large-scale sea water quality monitoring and management system is constructed. Disclosure of Invention In view of the above, the invention provides a large-scale sea area water pollution tracing monitoring site optimizing layout method for coupling efficient optimization and fine simulation, which approximates a water pollution transmission model through a radial basis function proxy model, applies Latin hypercube experimental design to pollution source identification, realizes efficient water pollution tracing, and further comprehensively considers cost and monitoring precision, and optimizes an optimizing scheme of a monitoring site. For this purpose, the invention provides the following technical scheme: A large-scale sea water pollution tracing monitoring site optimization layout method for coupling efficient optimization and fine simulation comprises the following steps: Adopting numerical simulation software to construct a three-dimensional hydrodynamic force-water quality coupling model; determining key decision variables of the three-dimensional hydrodynamic-water quality coupling model through Sobol sensitivity analysis; Initializing a radial basis function proxy model; car