Nonlinear synchronous fluorescence path selection method and device based on intelligent algorithm

Abstract

The invention discloses a nonlinear synchronous fluorescence path selection method and device based on an intelligent algorithm, and the method comprises the following steps: building a path optimization model with a genetic algorithm as a core based on a nonlinear synchronous fluorescence scanning principle, and setting a fitness function to represent the optimization planning of the path optimization model; and inputting the three-dimensional fluorescence spectrum data of a plurality of objects to be detected into the constructed path optimization model to obtain an optimal nonlinear synchronous scanning path. According to the method, the path screening time can be remarkably shortened, and the path quality and the analysis efficiency are improved, so that a more accurate and efficient non-linear synchronous fluorescence scanning path selection means is provided for researchers.

Description

Technical Field

[0001] The present invention relates to the field of fluorescence detection of multi-component substances, and particularly to a non-linear synchronous fluorescence path selection method and device based on an intelligent algorithm. Background Art

[0002] The non-linear synchronous fluorescence method is a technique that simultaneously scans along a certain curve passing through the contour spectral diagram. During the scanning process, the scanning rates of two monochromators can vary dynamically without maintaining a fixed ratio. This method can select corresponding paths according to different needs. By choosing an appropriate path, the influence of scattered light can be avoided in detection, the resolution of the spectral overlapping system can be further improved, the number of spectral scans can be reduced, and it is widely applied in the analysis of mixed component systems.

[0003] In the application of the non-linear synchronous fluorescence method, the most important point is the selection of the scanning path before measurement, and the selection basis is to obtain the best non-linear synchronous fluorescence spectrum. Generally speaking, this spectrum should have good separation effects on various analytes, with the maximum spectral intensity and minimum interference for each analyte.

[0004] However, in the selection of the non-linear synchronous fluorescence scanning path, it is generally determined by researchers through multiple attempts at present. That is, researchers first determine multiple scanning paths based on their own experience, and then obtain a series of fluorescence spectra scanned along these paths through actual detection, and select the path corresponding to the best-performing fluorescence spectrum as the scanning path. This method of path selection is not only time-consuming and laborious, but also highly dependent on the personal experience of researchers. To reduce the time consumed by researchers in the process of path selection, it is of great significance to develop a method that can quickly find a suitable scanning path. Summary of the Invention

[0005] The main object of the present invention is to overcome the path selection problem in the prior art, and propose a non-linear synchronous fluorescence path selection method and device based on an intelligent algorithm to help researchers quickly find a suitable scanning path and avoid the manpower and material resources required for a large number of experiments.

[0006] The present invention adopts the following technical solutions:

[0007] A non-linear synchronous fluorescence path selection method based on an intelligent algorithm includes the following:

[0008] Build a path optimization model with a genetic algorithm as the core based on the non-linear synchronous fluorescence scanning principle, and set a fitness function to characterize the optimization plan of the path optimization model; input the three-dimensional fluorescence spectral data of multiple analytes into the constructed path optimization model to obtain the optimal non-linear synchronous scanning path.

[0009] An optimal non - linear synchronous scanning path is obtained through the path optimization model, which specifically includes the following steps:

[0010] 1) Generate a weight distribution based on three - dimensional fluorescence spectrum data, and generate a set of initial scanning paths as the initial population based on the weight distribution;

[0011] 2) Set a fitness function based on peak intensity, signal interference control, path smoothness, and density distribution to evaluate the quality of each of the scanning paths;

[0012] 3) Calculate the fitness value of each of the initial scanning paths according to the fitness function, and use the tournament selection method to select the scanning paths with high fitness values as parents for reproduction;

[0013] 4) Find the best individual from the reproduced population, copy it as an elite to the next generation, perform crossover and mutation operations on the remaining individuals, and recombine the path control points of the scanning paths corresponding to the parents;

[0014] 5) Recalculate the fitness value and update the population, and determine whether the number of iterations or the fitness value reaches the set value. If so, output the optimal non - linear synchronous scanning path; if not, go to step 3).

[0015] The expression of the fitness function is as follows:

[0016] Fit = P - λI - αC - βR - γS+δD;

[0017] Where, Fit is the fitness value, P is the average peak intensity, reflecting the ability of the scanning path to capture the fluorescence peaks of each analyte; I is the total interference, representing the signal interference between analytes at the peak; C is the curvature penalty, used to suppress the excessive bending of the scanning path; R is the loop penalty, preventing the ineffective looping of the scanning path in the high - reward area; S is the self - intersection penalty, preventing the unreasonable self - intersection of the scanning path; D is the density reward, encouraging the scanning path to pass through high - importance areas; λ, α, β, γ, δ are weight factors.

[0018] The calculation formula of the average peak intensity is as follows:

[0019]

[0020] Detect peaks in the fluorescence intensity sequence of analyte j and record the maximum peak as pks j , if all analytes have available peaks, then P is the maximum average peak intensity; N is the number of path points of the scanning path; M is the number of analytes.

[0021] For the set of peak positions {I of analyte jj,m}, check the normalized values ​​z of all other analytes k≠j at this peak position k (l j,m ), the total interference is defined as:

[0022]

[0023] Among them, m is the index of each peak position detected, indicating that each peak position that may be generated in the path under the same object to be tested j is traversed.

[0024] The curvature penalty is characterized by penalizing the overly curved scanning path to avoid path mutation. The specific formula is defined as: Calculate the vector angle for every three adjacent points on the scanning path, assuming and Then we have:

[0025]

[0026] Where (x i-1 ,y i-1 ), (x i ,y i )、(x i+1 ,y i+1 ) are the coordinates of three adjacent points on the scanning path, i = 1, ..., N, N is the number of path points on the scanning path, and ε is a small value to prevent the denominator from being zero.

[0027] The detour penalty is characterized by the phenomenon that the scanning path detours ineffectively, and its calculation formula is defined as follows:

[0028] Define the direction angle φ of the i-th segment of the scanning path i =arctan2(y i -y i-1 ,x i -x i-1 ), then:

[0029]

[0030] Among them, (x i-1 ,y i-1 ), (x i ,y i ) are the coordinates of two adjacent points on the i-th segment of the scanning path, N is the number of path points on the scanning path; φ i-1 Indicates the direction angle corresponding to the (i-1)th segment.

[0031] The self-intersection penalty is characterized by the situation where all non-adjacent line segments intersect, and its formula is specifically defined as:

[0032] Divide the scanning path into N-1 line segments L i=(x i -y i-1 ,y i -y i-1 ), define the line segment intersection indicator function I(L i ,L j ):

[0033]

[0034] Among them, I(L i ,L j ) is the line segment L i and line segment L j The intersection indicator function of i -y i-1 ,y i -y i-1 ) are the coordinates of two adjacent points on the i-th segment on the scanning path.

[0035] The density reward is characterized by the average density of the area passed by the scanning path, and its formula is specifically defined as:

[0036]

[0037] Among them, (x i ,y i ) is the coordinate of the point on the i-th segment of the scanning path, D(x i ,y i ) is the average density of the area passed by the scanning path, and N is the number of path points.

[0038] A nonlinear synchronous fluorescence path selection device based on an intelligent algorithm, comprising:

[0039] A path optimization module, based on the principle of nonlinear synchronous fluorescence scanning and taking genetic algorithm as the core, sets a fitness function to characterize the optimization planning of the path optimization model;

[0040] The scanning path acquisition module inputs the three-dimensional fluorescence spectrum data of multiple objects to be tested into the constructed path optimization model to obtain the optimal nonlinear synchronous scanning path.

[0041] It can be seen from the above description of the present invention that, compared with the prior art, the present invention has the following beneficial effects:

[0042] The present invention makes full use of the synergistic advantages of computer technology and intelligent algorithms, transforms the spectroscopy principle into a mathematical model, and implements it through computer language. Combining nonlinear synchronous fluorescence technology with intelligent optimization algorithms can efficiently screen the optimal path, greatly reducing the time and energy spent by researchers in path selection, and improving path quality and analysis efficiency.

[0043] The present invention provides a systematic and automated optimization solution for the selection process of non-linear paths, which can solve the problems of time-consuming and laborious traditional methods and has effective practical value. At the same time, the present invention introduces spectral simulation technology, and provides reliable decision-making references for users through the theoretical spectra generated based on the selected paths, thereby effectively avoiding a large number of invalid experiments and significantly improving the experimental efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a flowchart of the path optimization model of the present invention.

[0045] Figure 2 It is a density map generated by the method proposed by the present invention.

[0046] Figure 3 It is a three-dimensional fluorescence contour map of benzo(a)pyrene (BaP) and fluoranthene (Fla) and the path selected by the method of the present invention.

[0047] Figure 4 It is a theoretical spectral simulation diagram of each component that may be obtained by scanning along the scanning path obtained by the method of the present invention.

[0048] Figure 5 It is the actual scanning spectrum of the scanning path obtained by the present invention.

[0049] The present invention will be further described in detail below in conjunction with the drawings and specific embodiments. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0050] The present invention will be further described below through specific embodiments.

[0051] Refer to Figure 1 , a non-linear synchronous fluorescence path selection method based on an intelligent algorithm, including the following: building a path optimization model with a genetic algorithm as the core based on the non-linear synchronous fluorescence scanning principle, and setting a fitness function to characterize the optimization plan of the path optimization model; inputting the three-dimensional fluorescence spectral data of multiple analytes into the constructed path optimization model to obtain the optimal non-linear synchronous scanning path.

[0052] The path selection principle proposed by the present invention refers to interpolating and automatically drawing spectral diagrams for the original data based on the three-dimensional fluorescence data of multiple analytes, and obtaining the fluorescence distribution of the analytes in a specific spectral range. It includes drawing a basic spectral diagram for scanning path selection, such as a contour spectral diagram, based on the three-dimensional fluorescence spectral data of each analyte, and further generating a density map based on the normalized fluorescence intensity and gradient information, and generating a scanning path passing through the high-signal region using the density map of the target region. The initial parameters of the intelligent optimization algorithm are determined by the substance objects to be detected. The present invention builds a path optimization model with a genetic algorithm as the core according to the non-linear synchronous fluorescence scanning principle.

[0053] The selection of the non - linear synchronous scanning path is synergistically affected by multiple factors. The present invention sets the adaptive adjustment of parameters, that is, a fitness function is constructed by comprehensively considering factors such as the peak intensity of multiple analytes, signal interference control, path smoothness, and density distribution to characterize the optimization plan.

[0054] The optimization model adopted by the present invention combines the optimization technology with the core of the adaptive intelligent evolutionary algorithm constructed based on the principle of non - linear synchronous fluorescence, and optimally selects the fitness function, so that when the operation path optimization model is run, an optimal non - linear synchronous scanning path that can effectively distinguish the fluorescence signals of each component, reduce signal overlap and interference can be obtained.

[0055] In the present invention, the three - dimensional fluorescence spectrum data of each analyte is the data that can be obtained by performing three - dimensional spectrum scanning with a general fluorescence instrument, which is generally a data matrix composed of excitation wavelength, emission wavelength, and the corresponding fluorescence intensity. The non - linear synchronous fluorescence scanning path is a curve, and fluorescence peaks corresponding to each analyte can be obtained by scanning along this path.

[0056] The present invention obtains the optimal non - linear synchronous scanning path through a path optimization model, which specifically includes the following:

[0057] 1) Generate a weight distribution based on the three - dimensional fluorescence spectrum data, and generate a set of initial scanning paths as the initial population based on the weight distribution.

[0058] 2) Set a fitness function based on peak intensity, signal interference control, path smoothness, and density distribution to evaluate the quality of each scanning path.

[0059] 3) Calculate the fitness value of each initial scanning path according to the fitness function, and adopt the tournament selection method to select the scanning path with a high fitness value as the parent for reproduction.

[0060] 4) Find the best individual from the reproduced population, copy it as an elite to the next generation, perform crossover and mutation operations on the remaining individuals, and recombine the path control points of the corresponding scanning paths of the parents.

[0061] 5) Recalculate the fitness value and update the population, and judge whether the number of iterations or the fitness value reaches the set value. If so, output the optimal non - linear synchronous scanning path; if not, enter step 3).

[0062] In step 1), the density map can effectively characterize the relative importance of different regions by fusing the normalized fluorescence intensity and gradient information, and is directly used as the weight distribution through mathematical transformation. Specifically, the density map is obtained by mapping the analyte spectrum to a two - dimensional plane through a formula:

[0063]

[0064] In the formula, Density(x,y) represents the density value at the coordinate point (x,y), Z i (x,y) represents the original fluorescence intensity value at the coordinate point (x,y), max represents the maximum value of all points in the density map, min represents the minimum value of all points in the density map, and ε is a small quantity to prevent the denominator from being zero. represents the normalization result of the i-th group of data. This formula represents the statistics of the density region passed by the scanning path, and the method used is to sample the mapping of the path points on the density map. In the spectral space, the non-linear synchronous scanning path is defined by multiple control points, and these control points determine the shape and position of the path. Then the initial population generation includes path construction, density map-guided generation, population diversity maintenance, etc.

[0065] Among them, path construction includes:

[0066] Bézier curve path: By selecting the starting point, ending point and several intermediate control points, a smooth non-linear path is generated.

[0067] Linear path: Composed of control points connected by straight line segments, suitable for scanning simple regions.

[0068] Hybrid path: Combining the characteristics of Bézier curves and linear segments, used to adapt to complex spectral intervals.

[0069] Density map-guided generation includes: Using the density map of the target area to preferentially generate paths passing through high-signal areas, improving the quality of the solutions of the initial population.

[0070] Population diversity maintenance includes: Ensuring the diversity of the initial population in terms of path shape and distribution, and avoiding the optimization process from falling into local extrema.

[0071] Among steps 2) and 3), the fitness function is used to evaluate the quality of each path, and its design comprehensively considers the following factors:

[0072] Peak intensity of fluorescence signal: Maximize the intensity of the peak signal on the path, improving the signal capture ability of the target analyte.

[0073] Signal interference: Minimize the signal overlap between different analytes, reducing the impact of interference on the detection results.

[0074] Path smoothness: By setting a penalty term for curvature, encourage the generation of smooth paths and avoid path mutations.

[0075] Path detour and self-intersection penalty: Detect whether there are unreasonable detours or self-intersections in the path and give corresponding penalties.

[0076] Density Reward: Encourage the high-signal regions in the path coverage density map to improve the overall efficiency of the path.

[0077] Based on this, the mathematical expression of the fitness function of the present invention can be expressed by the following formula:

[0078] Fit = P - λI - αC - βR - γS + δD;

[0079] Where, Fit is the fitness value, P is the average peak intensity, reflecting the ability of the scanning path to capture the fluorescence peaks of each object to be measured; I is the total interference, indicating the signal interference between objects to be measured at the peak; C is the curvature penalty, used to suppress the excessive bending of the scanning path; R is the loop penalty, preventing the scanning path from making ineffective loops in the high-reward area; S is the self-intersection penalty, preventing the scanning path from having unreasonable self-intersection phenomena; D is the density reward, encouraging the scanning path to pass through high-importance areas; λ, α, β, γ, δ are weight factors.

[0080] Among them, the calculation formula of the average peak intensity is as follows:

[0081]

[0082] Detect peaks in the fluorescence intensity sequence of object to be measured j and record the maximum peak as pks j , if all objects to be measured have available peaks, then P is the average maximum peak intensity; N is the number of path points of the scanning path; M is the number of objects to be measured.

[0083] For the set of peak positions {I j,m} of object to be measured j, check the normalized value z k (l j,m ) of all other objects to be measured k≠j at this peak position, and the total interference is defined as:

[0084]

[0085] Among them, m is the index of each detected peak position, indicating traversing each possible peak position generated by the path under the same object to be measured j.

[0086] The curvature penalty is characterized by punishing the overly bent scanning path to avoid the occurrence of path mutations. The specific formula is defined as: calculate the vector angle for every three adjacent points on the scanning path. Let and Then there is:

[0087]

[0088] C is the curvature penalty; (x i-1 , y i-1 ), (xi , y i ), (x i+1 , y i+1 ), (x

[0089] ), (x

[0090] ), (x i = arctan2(y i - y i-1 , x i - x i-1 ), then we have:

[0091]

[0092] where R is the loop penalty; (x i-1 , y i-1 ), (x i , y i ) are the coordinates of two adjacent points on the i-th segment of the scanning path, and N is the number of path points of the scanning path; φ i-1 is the direction angle corresponding to the (i - 1)-th segment. This formula is mainly used to measure the cumulative difference of the direction angles of each segment in the scanning path.

[0093] The self-intersection penalty is characterized by the intersection of all non-adjacent line segment pairs, and its formula is specifically defined as:

[0094] Divide the scanning path into N - 1 line segments L i = (x i - y i-1 , y i - y i-1 ), and define the line segment intersection indicator function I(L i , L j ):

[0095]

[0096] where S is the self-intersection penalty, and I(L i , L j ) is the intersection indicator function of line segment L i and line segment L j ; (x i-1 , y i-1 ), (x i , y i ) are the coordinates of two adjacent points on the i-th segment of the scanning path.

[0097] ​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​The density reward is characterized by the average density of the area passed by the scanning path, and its formula is specifically defined as:

[0098]

[0099] where (x i , y i ) are the coordinates of the points on the i-th segment of the scanning path, D is the density reward, D(x i , y i ) is the average density of the area passed by the scanning path, and N represents the number of path points.

[0100] In step 4), selection, crossover, mutation, and elitism retention are included. For selection, the tournament selection mechanism is used to pick individuals with higher fitness from the population as parents and retain the key genes of the excellent paths. Crossover is to perform crossover operations on the control points of the parent paths: randomly select crossover points, exchange the control points of the two paths in segments to generate offspring paths; and dynamically adjust the positions of the crossover points to fully combine the characteristics of Bezier, linear, and hybrid paths. Mutation is to introduce random mutations to enhance population diversity: adjust the positions of the path control points and add small Gaussian noise; and use the density map for local optimization to preferentially guide the path to high-signal areas. Elitism retention is to retain the individual with the highest current fitness to ensure the continuation of the optimal solution during the optimization process.

[0101] In order to further improve the algorithm efficiency and adapt to different experimental requirements, the present invention adds the following adaptive mechanisms:

[0102] Dynamic adjustment of fitness: According to the fitness distribution of each generation, adaptively adjust the normalization strategy to ensure the stability of the optimization process.

[0103] Dynamic regulation of mutation probability: When the population diversity decreases, appropriately increase the mutation probability to avoid premature convergence. When the population is concentrated near the optimal solution, reduce the mutation probability to accelerate convergence.

[0104] User feedback integration: Regularly display the current optimal path and optimize the direction based on user scores. Use the control points of the high-score paths as templates for offspring generation to improve the quality of the population solutions.

[0105] Step 5) includes termination conditions and output of the optimal path. The optimization process terminates when one of the following conditions is met:

[0106] Condition 1: The number of iterations reaches the preset maximum limit value (maxIter).

[0107] Condition 2: The optimal fitness value has not been significantly improved for multiple consecutive generations, indicating that the algorithm has converged.

[0108] Finally, the algorithm outputs the scanning path with the highest fitness in the current population as the optimal path for non-linear synchronous fluorescence scanning.

[0109] Based on the obtained optimal scanning path, the present invention simulates and verifies the fluorescence signal to achieve the separation and rapid identification of the components of the analyte: according to the determined scanning path, the algorithm is used to simulate and control the simultaneous scanning of the excitation and emission monochromators, record the intensity and wavelength of the non-linear synchronous fluorescence spectrum, obtain the fluorescence signals of each analyte, and draw the corresponding theoretical fluorescence spectrum for the researcher to modify. By using the method of the present invention, the excitation wavelength and emission wavelength corresponding to the optimal path are obtained, and the fluorescence intensity corresponding to each measurement point in the path is extracted from the mixed spectrum and plotted as a time series map, which is theoretically similar to the actually scanned spectrum, facilitating the researcher to judge the quality of the scanning path.

[0110] Compared with the traditional path selection method, the method of the present invention can significantly shorten the path screening time, improve the path quality and analysis efficiency, thus providing a more accurate and efficient means for researchers to select the non-linear synchronous fluorescence scanning path.

[0111] Based on this, the present invention also proposes a non-linear synchronous fluorescence path selection device based on an intelligent algorithm, which is used to execute the above-mentioned non-linear synchronous fluorescence path selection method based on an intelligent algorithm, including: a path optimization module, based on the non-linear synchronous fluorescence scanning principle and with the genetic algorithm as the core, setting a fitness function to characterize the optimization plan of the path optimization model. A scanning path acquisition module, inputting the three-dimensional fluorescence spectrum data of multiple analytes into the constructed path optimization model to obtain the optimal non-linear synchronous scanning path. Among them, the path optimization module is used to execute the above-mentioned steps 1)-5).

[0112] Example: Path selection for the simultaneous detection of benzo(a)anthracene (BaP) and fluoranthene (Fla)

[0113] The three-dimensional fluorescence contour maps of each analyte drawn based on the mixture of benzo(a)anthracene (BaP) and fluoranthene (Fla) are as Figure 3 shown, and the generated density weight map is as Figure 2 shown, which is used to guide the generation of the initial population and the adjustment of the path. According to the separation requirements of the two substances, 30 iterations are set, and 400 path points are used as the initial parameters. In this embodiment, through the optimization technology with the adaptive genetic algorithm as the core, the fitness of each path is examined, the paths with high fitness are rewarded, and it is inclined to generate offspring similar to them, the paths with low fitness are punished, and the paths similar to them are reduced in the generated offspring. After 30 iterations, it is found that the optimal fitness value remains basically unchanged in the later stage of iteration. Therefore, it is considered that the path obtained by the algorithm meets the preset goal, and the selected path is asFigure 3 As shown. Based on Figure 3 the optimal path obtained in Figure 4 the corresponding mapping between the path points and the corresponding points in the three-dimensional spectrum, a simulated fluorescence spectrum diagram is obtained as shown in

[0114] After obtaining the scanning path (as shown in Appendix 1), it is imported into an instrument with non-linear synchronous fluorescence scanning function to carry out actual detection work. In this embodiment, the results of detection along the selected optimal scanning path are as shown in Figure 5 the fluorescence signals of the two analytes are effectively separated and are basically independent of each other. This result verifies the practicability and reliability of the present invention, and can effectively assist researchers in quickly screening out an optimal path suitable for non-linear synchronous fluorescence scanning, effectively improving the experimental efficiency.

[0115] Appendix 1 Example of Scanning Path

[0116]

[0117]

[0118] It should be noted that although several modules or units of the devices for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of the two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0119] Through the description of the above embodiments, those skilled in the art can easily understand that the example embodiments described herein can be implemented by software or by a combination of software and necessary hardware. Therefore, the technical solutions according to the embodiments of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on the network, including several instructions to enable a computing device (which can be a personal computer, a server, a touch terminal, or a network device, etc.) to execute the method according to the embodiments of the present disclosure.

[0120] After considering the specification and practicing the invention disclosed herein, those skilled in the art will readily conceive of other embodiments of the present disclosure. The present disclosure is intended to cover any variations, uses, or adaptations of the present disclosure, which follow the general principles of the present disclosure and include known common knowledge or conventional technical means in the technical field not disclosed by the present disclosure.

[0121] The above are only specific embodiments of the present invention, but the design concept of the present invention is not limited thereto. Any non-substantive modification of the present invention using this concept shall fall within the scope of infringement of the protection scope of the present invention.

Claims

1. A nonlinear synchronous fluorescence path selection method based on an intelligent algorithm, characterized in that: These include: Based on the principle of nonlinear synchronous fluorescence scanning, a path optimization model with genetic algorithm as the core is built, and a fitness function is set to characterize the optimization planning of the path optimization model; the three-dimensional fluorescence spectrum data of multiple objects to be tested are input into the constructed path optimization model to obtain the optimal nonlinear synchronous scanning path.

2. The nonlinear synchronous fluorescence path selection method based on intelligent algorithm according to claim 1, characterized in that: The optimal nonlinear synchronous scanning path is obtained by the path optimization model, which specifically includes the following: 1) generating a weight distribution based on the three-dimensional fluorescence spectrum data, and generating a set of initial scanning paths as an initial population based on the weight distribution; 2) Setting a fitness function based on peak intensity, signal interference control, path smoothness and density distribution to evaluate the quality of each scanning path; 3) calculating the fitness value of each of the initial scanning paths according to the fitness function, and selecting the scanning path with a high fitness value as the parent generation for reproduction by using a tournament selection method; 4) Find the best individual from the population after reproduction, copy it as the elite to the next generation, perform crossover and mutation operations on the remaining individuals, and reorganize the path control points of the scanning path corresponding to the parent generation; 5) Recalculate the fitness value and update the population to determine whether the number of iterations or the fitness value reaches the set value. If so, output the optimal nonlinear synchronous scanning path; if not, proceed to step 3).

3. The nonlinear synchronous fluorescence path selection method based on intelligent algorithm according to claim 2, characterized in that: The fitness function expression is as follows: Fit=P-λI-αC-βR-γS+δD; Among them, Fit is the adaptability value, P is the average peak intensity, which reflects the ability of the scanning path to capture the fluorescence peak of each analyte; I is the total interference, which indicates the signal interference between the analytes at the peak; C is the curvature penalty, which is used to suppress excessive bending of the scanning path; R is the circle penalty, which prevents the scanning path from invalidly circling in the high reward area; S is the self-intersection penalty, which prevents the scanning path from having unreasonable self-intersection; D is the density reward, which encourages the scanning path to pass through the high-importance area; λ, α, β, γ, δ are weight factors.

4. The nonlinear synchronous fluorescence path selection method based on intelligent algorithm according to claim 3, characterized in that: The calculation formula of the average peak intensity is as follows: The fluorescence intensity sequence of the analyte j Detect the peak value in the test, and record the maximum peak value as pks j , if all the objects to be tested have available peak values, then P is the maximum average peak intensity; N is the number of path points in the scanning path; and M is the number of objects to be tested.

5. The nonlinear synchronous fluorescence path selection method based on intelligent algorithm according to claim 3, characterized in that: For the peak position set {I j,m }, check the normalized values ​​z of all other analytes k≠j at this peak position k (l j,m ), the total interference is defined as: Among them, m is the index of each peak position detected, indicating that each peak position that may be generated in the path under the same object to be tested j is traversed.

6. The nonlinear synchronous fluorescence path selection method based on intelligent algorithm according to claim 3, characterized in that: The curvature penalty is characterized by penalizing the overly curved scanning path to avoid path mutation. The specific formula is defined as: Calculate the vector angle for every three adjacent points on the scanning path, assuming and Then we have: Where (x i-1 ,y i-1 ), (x i ,y i )、(x i+1 ,y i+1 ) are the coordinates of three adjacent points on the scanning path, i = 1, ..., N, N is the number of path points on the scanning path, and ε is a small value to prevent the denominator from being zero.

7. The nonlinear synchronous fluorescence path selection method based on intelligent algorithm according to claim 3, characterized in that: The detour penalty is characterized by the phenomenon that the scanning path detours ineffectively, and its calculation formula is defined as follows: Define the direction angle φ of the i-th segment of the scanning path i =arctan2(y i -y i-1 ,x i -x i-1 ), then: Among them, (x i-1 ,y i-1 ), (x i ,y i ) are the coordinates of two adjacent points on the i-th segment of the scanning path, N is the number of path points on the scanning path; φ i-1 Indicates the direction angle corresponding to the (i-1)th segment.

8. The nonlinear synchronous fluorescence path selection method based on intelligent algorithm according to claim 3, characterized in that: The self-intersection penalty is characterized by the situation where all non-adjacent line segments intersect, and its formula is specifically defined as: Divide the scanning path into N-1 line segments L i =(x i -y i-1 ,y i -y i-1 ), define the line segment intersection indicator function I(L i ,L j ): Among them, I(L i ,L j ) is the line segment L i and line segment L j The intersection indicator function of i-1 ,y i-1 ), (x i ,y i ) are the coordinates of two adjacent points on the i-th segment on the scanning path.

9. The nonlinear synchronous fluorescence path selection method based on intelligent algorithm according to claim 3, characterized in that: The density reward is characterized by the average density of the area passed by the scanning path, and its formula is specifically defined as: Among them, (x i ,y i ) is the coordinate of the point on the i-th segment of the scanning path, D(x i ,y i ) is the average density of the area passed by the scanning path, and N is the number of path points.

10. A nonlinear synchronous fluorescence path selection device based on intelligent algorithm, characterized in that: include: A path optimization module, based on the principle of nonlinear synchronous fluorescence scanning and taking genetic algorithm as the core, sets a fitness function to characterize the optimization planning of the path optimization model; The scanning path acquisition module inputs the three-dimensional fluorescence spectrum data of multiple objects to be tested into the constructed path optimization model to obtain the optimal nonlinear synchronous scanning path.

Images