CN113588572B - Intelligent management system for farmland heavy metal online detection correction model - Google Patents
Intelligent management system for farmland heavy metal online detection correction model Download PDFInfo
- Publication number
- CN113588572B CN113588572B CN202110892464.8A CN202110892464A CN113588572B CN 113588572 B CN113588572 B CN 113588572B CN 202110892464 A CN202110892464 A CN 202110892464A CN 113588572 B CN113588572 B CN 113588572B
- Authority
- CN
- China
- Prior art keywords
- correction
- model
- index
- spectrum
- heavy metal
- 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.)
- Active
Links
- 238000012937 correction Methods 0.000 title claims abstract description 291
- 229910001385 heavy metal Inorganic materials 0.000 title claims abstract description 78
- 238000001514 detection method Methods 0.000 title claims abstract description 69
- 238000001228 spectrum Methods 0.000 claims abstract description 125
- 238000011156 evaluation Methods 0.000 claims abstract description 87
- 238000000034 method Methods 0.000 claims abstract description 45
- 239000002689 soil Substances 0.000 claims abstract description 26
- 230000000694 effects Effects 0.000 claims abstract description 12
- 238000004364 calculation method Methods 0.000 claims abstract description 8
- 238000004458 analytical method Methods 0.000 claims description 13
- 238000003860 storage Methods 0.000 claims description 5
- 238000002474 experimental method Methods 0.000 claims description 3
- 230000003044 adaptive effect Effects 0.000 abstract description 3
- 238000012545 processing Methods 0.000 abstract description 3
- 230000003595 spectral effect Effects 0.000 description 15
- 230000009466 transformation Effects 0.000 description 14
- 238000013461 design Methods 0.000 description 6
- 238000012827 research and development Methods 0.000 description 5
- 238000011161 development Methods 0.000 description 4
- 238000000354 decomposition reaction Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000010183 spectrum analysis Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000004445 quantitative analysis Methods 0.000 description 2
- 238000012742 biochemical analysis Methods 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000004587 chromatography analysis Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000002790 cross-validation Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000000840 electrochemical analysis Methods 0.000 description 1
- 235000013305 food Nutrition 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000004949 mass spectrometry Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 229910052751 metal Inorganic materials 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 238000004611 spectroscopical analysis Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 208000011580 syndromic disease Diseases 0.000 description 1
- 238000001308 synthesis method Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000002076 thermal analysis method Methods 0.000 description 1
- 230000001131 transforming effect Effects 0.000 description 1
- 235000013311 vegetables Nutrition 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/27—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands using photo-electric detection ; circuits for computing concentration
- G01N21/274—Calibration, base line adjustment, drift correction
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Theoretical Computer Science (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention discloses an intelligent management system for an online detection and correction model of farmland heavy metals, which comprises the following components: the model registration module is used for providing registration service for correction models of various heavy metal detectors; the spectrum input module is used for providing input interface services of original spectrum data of the soil standard sample and original spectrum data of the soil detection sample; the index selection module is used for selecting index sub-services of the correction model and deriving different correction models through different index sub-service combinations; the data correction module is used for carrying out data correction calculation on the original spectrum; the correction evaluation module is used for evaluating the correction effect of the standard sample after the original spectrum correction, and a model with the optimal evaluation result is used as a correction model corresponding to the heavy metal detector; and the spectrum output module is used for providing a corrected spectrum data output service. When the instrument performs correction model selection, an optimal method or model can be called to perform spectrum correction processing, and adaptive correction service is provided.
Description
Technical Field
The invention belongs to the technical field of farmland ecosystems, relates to the technical field of soil heavy metal detection and correction models, and particularly relates to an intelligent management system of an online detection and correction model of farmland heavy metals.
Background
At present, with the deep development and utilization of natural resources, the increase of food and grain safety requirements and other factors, the land is taken as the most basic strategic resource, and is always the key point of protection and research. The heavy metal content in the farmland directly affects the farmland quality, and the heavy metals such as grain crops, vegetables and fruits planted in the land exceeding the national standard of the heavy metal content have a high probability of exceeding the standard, and influence the human health. Therefore, accurate quantitative analysis of farmland heavy metals has been the research direction of heavy metal analyzers.
Accurate quantitative analysis mainly relies on standard sampling and accurate detection after sample preparation, and detection needs detection principle support. At present, heavy metal detection equipment mainly comprises: the main analysis principles of the large-scale and medium-scale heavy metal analyzers and the handheld small-scale heavy metal analyzers in laboratories relate to various analysis methods such as electrochemical analysis, spectral analysis, chromatographic analysis, mass spectrometry, spectral analysis, thermal analysis, radiation analysis, biochemical analysis, chemical analysis and the like. And with the rapid increase of quantitative determination tasks and sample quantity, the timeliness requirement of measurement is improved, and the small heavy metal spectrum analysis equipment has the advantages of rapid detection in farmlands, short-time detection of multiple points, portability and the like, and rapidly becomes a research hotspot.
However, the existing metal element spectrum detection equipment in industry cannot be directly used for detection due to various factors such as soil matrix, water content and the like of farmland soil, and heavy metal detection equipment meeting the field detection requirements needs to be designed, developed and manufactured aiming at farmland environment. In particular to heavy metal on-site rapid detection equipment which only needs simple soil pretreatment and even does not need soil pretreatment. Because of a plurality of interference factors, in order to ensure the accuracy and repeatability of the detection result, not only each parameter of the light path needs to be reasonably designed during design, but also the spectral correction by software at the later stage is very important. Therefore, spectral correction becomes an important ring, and the acquisition of accurate detection data requires timely correction of the original measurement data.
The development of correction systems has an important role in spectroscopic analysis. However, the existing correction exists mainly in the following ways: the correction of the instrument is fused in the whole system of the heavy metal spectrometer, is tightly coupled with the spectrometer, and is carried out by adopting a single correction method. The correction implementation principle mainly comprises the following steps: the tight coupling correction is performed from the aspect of combining hardware and software, and the instrument uses an integrated correction system design. The method has the defects of high correction cost, insufficient plug-in design of the correction system, difficult formation of plug-in correction design and insufficient correction evaluation.
In summary, the existing mode and implementation principle of the correction are combined, the correction analysis of the conventional farmland heavy metal analyzer is mainly in the equipment informatization stage, the integrated design research and development of hardware and software causes the closed loop circulation of a software and hardware correction method, and the processing accuracy degree depends on the research and development level of enterprises, so that the credibility of production data is bound with an instrument.
Therefore, the following problems exist in summary of the prior art:
(1) The correction method is single, and the correction system has insufficient adaptability.
The integrated design of the software and hardware of the farmland heavy metal instrument causes the closed loop circulation of the correction method, the method or system parameter is single, only one instrument is adapted, or only one instrument can be adapted in the enterprise, the improvement degree of the correction method depends on the research and development level and research and development condition of the enterprise, an open farmland heavy metal analysis and correction ecosystem is not formed, namely, the on-line intelligent detection correction model management cannot be carried out according to the adjustment of the correction combination of the heavy metal detection equipment, and the adaptive correction service combination cannot be provided.
(2) The evaluation method is single, and the comprehensive evaluation index is not formed.
The evaluation method and the evaluation index after the spectrum correction of the farmland heavy metal detector used by each enterprise are different, the evaluation result has no universality, the problem that the evaluation effect is good by using a certain method and the evaluation effect is poor by using another method is easily caused, and a complete comprehensive evaluation index model is formed by not integrating various evaluation methods.
Disclosure of Invention
In view of the defects of the prior art, the invention provides an intelligent management system for an online detection and correction model of farmland heavy metal, which is mainly based on the rapid development trend of the information in the current Internet age, and the information of equipment is improved to the requirement of industry information in the research and development and utilization processes of farmland heavy metal detection instruments. From the informatization perspective of the farmland heavy metal determination industry, an online intelligent detection correction model management system for farmland heavy metals needs to be established, and an open and evaluable management system for intelligent detection correction models is formed.
In order to achieve the above purpose, the technical scheme adopted by the invention is as follows:
the embodiment of the invention provides an intelligent management system for an online detection and correction model of farmland heavy metals, which comprises the following components:
the model registration module is used for providing registration service for correction models of various heavy metal detectors;
the spectrum input module is used for providing input interface services of original spectrum data of the soil standard sample and original spectrum data of the soil detection sample;
the index selection module is used for selecting index sub-services of the correction model and deriving different correction models through different index sub-service combinations;
the data correction module is used for carrying out data correction calculation on the original spectrum;
the correction evaluation module is used for evaluating the correction effect of the standard sample after the original spectrum correction, and a model with the optimal evaluation result is used as a correction model corresponding to the heavy metal detector;
the spectrum output module is used for providing a corrected spectrum data output service; the detection sample correction spectrum data are directly stored into a correction spectrum database for subsequent analysis.
Further, the model registration module includes:
the description specification unit is used for uniformly describing model indexes of correction models of various heavy metal detectors by adopting a structural modeling language and establishing a model description specification;
the standardized registration unit is used for realizing the full standardized registration of the existing correction model by identifying the standardized description file of the description specification unit;
and the storage unit is used for uniformly storing the correction models registered by the standardized registration unit in a correction model database and storing the correction models according to a classification table of the correction models.
Further, the spectrum input module comprises:
the forming stage unit is used for calibrating each parameter and parameter value of the model by using the original spectrum data of the soil standard sample in the correcting model forming stage, and the calibrated correcting model can be stored in a correcting model database;
an application stage unit for using the original spectrum data of the soil detection sample in a correction model application stage and calling a correction model in the correction model database to finish correction of the original spectrum information;
the original spectrum data of the soil standard sample and the original spectrum data of the soil detection sample which need to be corrected are provided with the equipment numbers of the instrument and are used for being uniquely matched with the correction model.
Further, the index selection module includes:
the standard sample selection unit is used for when a certain heavy metal detector does not have a corresponding correction model; the index selection module uses the input standard sample original spectrum data to traverse different methods corresponding to each standard correction model preset in the system, and selects corresponding required index sub-services from different methods, wherein the same index has different parameters and parameter values for selection;
the detection sample selection unit is used for matching the correction model in the correction model library according to the equipment number in the input detection sample original spectrum data when a certain heavy metal detector has a corresponding correction model, and then selecting model indexes and parameter values according to the index sub-service appointed during the registration of the correction model; if other experiments prove that the model index value has better choice, the standardized registration unit adjusts the original correction model index and the index value thereof, and uses the new correction model to correct.
Further, the data correction module includes:
the first data correction unit is used for correcting and calculating the original spectrum data according to the index sub-service combination selected by the index selection module and outputting the corrected data to the correction evaluation module;
the second data correction unit is used for correcting and calculating the original spectrum data according to the index sub-service combination selected by the index selection module and outputting the corrected data to the spectrum output service.
Further, the correction evaluation module includes:
the evaluation threshold unit is used for setting comprehensive correction evaluation indexes and thresholds, measuring correction evaluation results and thresholds, and if the threshold conditions are met, selecting system automation evaluation to obtain correction model results; if the evaluation result does not meet the threshold condition, repeating the index selection, the data correction and the correction evaluation until the evaluation result reaches the threshold condition;
and the comprehensive evaluation unit evaluates the correction result of each service combination of the spectrum of the standard sample by utilizing the comprehensive evaluation index and the threshold set by the evaluation threshold unit, and the correction model of the optimal evaluation result is used as the correction model of the heavy metal detection equipment and is registered by using the model registration module.
Further, the spectral output module comprises:
a spectrum output unit for outputting corrected spectrum data;
and a storage unit storing the corrected spectrum data for subsequent analysis.
Compared with the prior art, the invention has the following beneficial effects:
the intelligent management system for the farmland heavy metal online detection correction model provided by the embodiment of the invention comprises the following components:
(1) The online intelligent detection correction model management system can enhance the management of the original system on the correction method or model. After the original system is integrated with the correction model intelligent management system provided by the invention, an open farmland heavy metal analysis and correction ecological system can be formed. When the instrument performs correction model selection, a corresponding method or model can be found according to the related parameters of the method or model, and an optimal method or model is selected according to the evaluation index; when the spectrum correction is carried out, the selected optimal method or model can be called to carry out spectrum correction processing, and adaptive correction service is provided.
(2) The online intelligent detection correction model management system can manage correction models of various heavy metal detectors online and take on tasks such as correction of original spectrum data of the heavy metal detectors. The specific management content comprises the registration, storage, index adjustment, optimal correction method or model evaluation and selection and the like of the correction models of the heavy metal detectors of different types. For different types of heavy metal detectors, the correction effect evaluation service can select an optimal correction model for the heavy metal detectors to form a comprehensive evaluation index. Namely, each type of heavy metal detector subjected to correction and evaluation corresponds to only one correction model. After the correction model is selected, the corresponding correction model can be directly indexed by subsequent spectral correction.
(3) The integration of the heavy metal detector to the intelligent detection correction model management system is beneficial to the decoupling of the software and hardware of the farmland heavy metal instrument, the expansion of a correction library, the accuracy evaluation of a correction method and the consistency evaluation of correction results.
Drawings
Fig. 1 is a service combination flow chart of an intelligent management system for an online detection and correction model of farmland heavy metals, which is provided by the embodiment of the invention.
Detailed Description
The invention is further described in connection with the following detailed description, in order to make the technical means, the creation characteristics, the achievement of the purpose and the effect of the invention easy to understand.
In the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "upper", "lower", "inner", "outer", "front", "rear", "both ends", "one end", "the other end", etc. are based on the directions or positional relationships shown in the drawings, are merely for convenience of describing the present invention and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a specific direction, be configured and operated in the specific direction, and thus should not be construed as limiting the present invention. Furthermore, the terms "first," "second," and the like, are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
In the description of the present invention, it should be noted that, unless explicitly specified and limited otherwise, the terms "mounted," "provided," "connected," and the like are to be construed broadly, and may be fixedly connected, detachably connected, or integrally connected, for example; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium, and can be communication between two elements. The specific meaning of the above terms in the present invention will be understood in specific cases by those of ordinary skill in the art.
The intelligent management system for the farmland heavy metal online detection correction model provided by the embodiment of the invention is used for model management and data correction, and is realized by providing a correction service combination model online or downloading a dynamic link library of the correction service combination model online and using the dynamic link library locally through implicit link or explicit link. The dynamic link library contains multiple formats such as dll and so, and can be operated in multiple systems such as Windows, linux, android.
Referring to fig. 1, the method includes:
the model registration module is used for providing registration service for correction models of various heavy metal detectors;
the spectrum input module is used for providing input interface services of original spectrum data of the soil standard sample and original spectrum data of the soil detection sample;
the index selection module is used for selecting index sub-services of the correction model and deriving different correction models through different index sub-service combinations;
the data correction module is used for carrying out data correction calculation on the original spectrum;
the correction evaluation module is used for evaluating the correction effect of the standard sample after the original spectrum correction, and a model with the optimal evaluation result is used as a correction model corresponding to the heavy metal detector;
the spectrum output module is used for providing a corrected spectrum data output service; the detection sample correction spectrum data are directly stored into a correction spectrum database for subsequent analysis.
The correction service composition model structure is described below:
the correction services are combined into a tree structure, and the overall service is described as S k ＝{m k ,spi k ,i k ,s k ,ef k ,spo k -comprising model registration m k Spectral input sci k Index selection i k Data correction s k Correction and evaluation of ef k Spectral output spo k Six classes of service. Model registration, spectrum input without front service, spectrum output without back service, and in addition, each class of service has front service and back service, the input of front service is used as the input of the service, the output of the service is used as the input of the back service, each class of service comprises a plurality of selectable sub-services, and different sub-services are provided when different requirements exist. Each service in the six classes of services of the registration model is different in service combination because of the fact that the services of each class relate to different numbers, different types and different parameters of sub-services, and each service combination is finally stored in a data table of the same class model as a service combination record. Wherein each service can be described as w k ＝{s pre ,in k ,out k ,s post }，s pre Is a front-end service, in is a service input, out k S for service output post Serving the post-consumer.
The relation among the services in the correction service combination is as follows: the original spectrum data is input through a model input service, after the input, the equipment type numbers of the instrument are used for matching correction models in a correction model library, if no corresponding correction model exists, standard sample original spectrum is needed to be input, different methods corresponding to preset standard correction models are traversed through an index selection service, the needed index sub-service is selected from the different methods, and the same index has different parameters and parameter values for selection. And then, calling a data correction service to complete the correction of the original spectrum data, evaluating the original spectrum data by using a correction evaluation service, and feeding back the result to an index selection service. The correction model with the optimal evaluation result is used as the correction model of the heavy metal detection equipment, the model registration is carried out in the model registration service, and finally the correction model is stored in a correction model library. If the corresponding correction model is available, the input original spectrum is treated by the detection sample spectrum, the corresponding correction model in the correction model library is matched, the index selection service selects according to the index sub-service selected by the model, the subsequent data correction service is completed, and the spectrum output service is used for outputting.
(1) Model registration service: the purpose is to provide registration service for correction models of various heavy metal detectors (laboratory large-scale and medium-scale heavy metal analyzers and handheld small-scale heavy metal analyzers). The correction system uniformly uses structured modeling languages such as XML, JSON and the like to describe model indexes of various models, and establishes model description specifications. The system realizes the full-scale standardized registration of the existing correction model through the identification of the standardized description file. After registration, the models are uniformly stored in a correction model database, and are stored according to different types of correction models and different data tables, the same type of models are stored in the same piece of data table, and each record in the data table represents a model with different index values.
Model registration is largely divided into two forms: the first is user registration, the user directly inputs model parameters and parameter values according to standard specifications of model registration, the model is manually registered, and the correction model is used for spectrum correction of the subsequent heavy metal detector. The second is system registration, when the model and the specific parameters and parameter values of the model can not be accurately known, the system registers the correction model meeting the set correction requirement after index selection, data correction, correction evaluation and other services are carried out by using the spectrum of the standard sample, and the correction model is stored in a correction model library.
(2) Spectral input service: for providing the original spectral numberAccording to the input interface service. The input original spectrum data comprises original spectrum data of a soil standard sample and original spectrum data of a soil detection sample, and the input services are spi respectively k1 、spi k2 . Wherein, the soil original spectrum data stream to be corrected needs to carry the equipment type number of the instrument. The system calls different spectrum input services mainly depending on whether the equipment type number of the heavy metal instrument has a unique corresponding correction model in the correction model library. Wherein, original spectrum data of soil standard sample is input into service spi k1 The calibration method is used for calibrating parameters and parameter values of a calibration model without matching of the heavy metal detector, namely, in a calibration model forming stage; soil detection sample original spectrum data input service spi k2 The method is used for a heavy metal detector and is provided with a matched correction model, namely, in the application stage of the correction model, an input spectrum is treated as a detection sample spectrum in spectrum input service no matter the input spectrum is a standard sample spectrum or a detection sample spectrum.
(3) Index selection service: the method is used for selecting index sub-services of the correction model and deriving different correction models through different index sub-service combinations. Different types of original spectrum data are input, correction indexes corresponding to the correction models are different, and different sub-services are required to be called when the models are called.
The spectrum input can uniquely match the correction models in the correction model library through the device type field carried by the spectrum. When a certain heavy metal detector does not have a corresponding correction model, the working principle of the index selection service is as follows: the index selection service uses the input standard sample original spectrum data, traverses different methods corresponding to each standard correction model prefabricated in the system, selects the required index sub-service from the different methods, and the same index has different parameters and parameter values for selection.
When a certain heavy metal detector has a corresponding correction model, the working principle of the index selection service is as follows: the input original spectrum is matched with a correction model in a correction model library, and the index selection service selects according to index sub-services selected by the model. In special cases, if other experiments prove that the model index sub-service has better choice, the original correction model index and the index value thereof can be adjusted by utilizing the man-machine interaction page, and the correction is performed by using the new correction model.
Taking a wavelet transformation model as an example, the model index is described as follows:
wavelet transformation includes continuous transformation, discrete transformation, fast transformation, smooth transformation, lifting transformation, and other transformation modes. Which uses the purpose of wavelet transformation to correct the spectral signal, i.e. noise removal. The wavelet denoising method comprises wavelet transformation mode maximum denoising, wavelet transformation threshold denoising, wavelet coefficient correlation denoising, multi-wavelet denoising, projection principle-based matching pursuit denoising, wavelet coefficient model denoising and synthesis method denoising (wavelet transformation+classical method), and the like. Among them, the most common are wavelet transform mode maximum denoising, wavelet transform threshold denoising, wavelet coefficient correlation denoising. Therefore, the specification will be described in detail with the index selection of three types of denoising methods.
The wavelet transformation model is mainly divided into four indexes, namely wavelet base i k1 Wavelet decomposition level i k2 Denoising mode i k3 Evaluation index i k4 . The wavelet transformation threshold denoising index decomposition is more detailed, and is wavelet basis, wavelet decomposition level, threshold processing method, threshold function type and evaluation index. The wavelet base selection of the spectrum signal is mainly determined by four aspects of requirements of support length, symmetry, vanishing moment characteristic and regularity of the wavelet function.
And selecting a correction model index suitable for a certain type of heavy metal detector by transforming the combination of different index sub-services and combining the evaluation index.
Table-one wavelet denoising index table
Remarks: 1) Wavelet basis, child wavelet (mother wavelet is scaled and translated, conservation of energy).
2) Wavelet transform mode maximum denoising: when the number of the interference pulse points is small, namely omega is small, the denoising effect of the threshold method is better than that of the mode maximum value method; when the noise is increased, the mode maximum value method denoising is obviously superior to the threshold denoising; but as ω continues to increase, the effect of mode maximum denoising is also better than threshold denoising (ω=the number of points of impulse noise/the number of points of the original signal).
3) Wavelet coefficient correlation denoising: and the correlation among wavelet coefficient scales is used for taking the coefficients out according to the correlation of the wavelet coefficients of adjacent scales, so that the real signal is reconstructed.
(4) Data correction service: a method for performing data correction calculations on an original spectrum, comprising: and carrying out data correction calculation on the original spectrum of the standard sample and carrying out data correction calculation on the original spectrum of the detection sample. The principle is as follows: the data correction service performs correction calculation on the original spectrum data according to the index sub-service combination selected by the index selection service. The data correction service comprises a plurality of sub-services, one for each type of correction model. When using the correction model, the data correction service provides a class of correction model sub-services s to the original spectrum k . If the spectrum is the original spectrum of the standard sample, outputting the corrected spectrum to a correction evaluation service; and if the spectrum is the original spectrum of the detection sample, outputting the corrected spectrum to a spectrum output service.
(5) Correction evaluation service: the method is used for evaluating the correction effect of the standard sample original spectrum after correction, a model with the optimal evaluation result is used as a correction model corresponding to the heavy metal detector, and correction evaluation service is not needed for performing spectrum correction on the detection sample original spectrum. The principle is as follows: after the correction evaluation service acquires the spectrum data output by the data correction service, the correction result is evaluated through the comprehensive evaluation index, and the correction model of the optimal evaluation result is used as the correction model of the heavy metal detection equipment.
Setting a threshold value for the evaluation result of the comprehensive evaluation index, if the evaluation result meets the threshold value condition, selecting a correction model result obtained by system automation evaluation, and if the evaluation result does not meet the threshold value condition, repeating the index selection, the data correction and the correction evaluation until the evaluation result reaches the threshold value condition.
The evaluation index comprises a plurality of dimensions, namely correction evaluation of the standard sample spectrum one-dimensional signal, such as signal to noise ratio, smoothness and the like; and the method also comprises correction evaluation of the heavy metal content value of the standard sample detected after correction relative to a true value, such as root mean square error and the like.
Table two evaluation methods table
Combining all indexes, wherein part of indexes have the same meaning and different names and are listed as different index items:
classification:
1) Signal-to-noise ratio, signal-to-noise ratio gain;
2) Smoothness;
3) Information entropy;
4) Correcting the decision coefficient, verifying the decision coefficient, fitting the coefficient and the correlation coefficient;
5) Mean square error, root mean square error, internal cross validation mean square error, corrected standard deviation, predicted root mean square error, predicted standard deviation, relative prediction error, relative error average;
6) Peak relative error;
currently, one class is selected from each class as a representative entry formula, and a comprehensive index θ is established, for example:
the larger θ, the better the noise removal result. Under the condition that a certain index is missing, the index defaults to 1, and the evaluation result is not affected.
(6) Spectral output service: for providing a corrected spectral data output service. Wherein the test sample correction spectral data is directly stored into the correction spectral database for subsequent analysis.
In summary, the farmland heavy metal online detection correction model intelligent management system provided by the embodiment of the invention forms the industry informatization of the correction model intelligent management system from the aspects of a method and a model, and mainly realizes two aspects: and correcting the service combination model and the comprehensive evaluation index.
(1) Correction service composition model
Correction service composition model: and providing combinable correction services, combining different correction models through different syndrome services, and automatically correcting the original spectrum information. Model includes model registration m k Spectral input sci k Index selection i k Data correction s k Correction and evaluation of ef k Spectral output spo k Six major classes of services and corresponding sub-services. The original spectrum comprises an original spectrum of a standard sample and a detection sample, and the original spectrum is operated through different business processes respectively. The original spectrum of the standard sample is corrected by adopting a registered standard correction model, the index selection is carried out according to the index fixed by the standard correction model, and the correction index parameter and index value are adjusted according to the multiple correction effect evaluation until the set correction target is reached. And the original spectrum of the detection sample is subjected to spectrum correction by adopting an existing corresponding model. The correction model library belongs to an open database, all developers using the correction model database can submit correction methods or models, unique correction methods or models are provided, and the diversity and adaptability of the correction methods and the model library are improved.
(2) Comprehensive evaluation index
And comprehensively selecting different types of correction indexes to form a comprehensive evaluation index theta, wherein the comprehensive evaluation index is divided into two parts, one part is used for evaluating the correction signals, and the other part is used for evaluating the correction values. The comprehensive evaluation index can evaluate the correction model according to the evaluation item, so that the correction reliability of the correction model is enhanced.
The intelligent management system for the farmland heavy metal online detection correction model provided by the embodiment of the invention realizes the functions of registration, index selection, index value adjustment, model open sharing and the like of a correction method or model, and feeds back the correction selection method and/or model through comprehensive evaluation indexes, so that the interconnection and intercommunication of the farmland heavy metal detection industry in the field of data correction and evaluation are further promoted, and the development from equipment informatization to industry informatization of farmland heavy metal detection is promoted.
The foregoing has shown and described the basic principles and main features of the present invention and the advantages of the present invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, and that the above embodiments and descriptions are merely illustrative of the principles of the present invention, and various changes and modifications may be made without departing from the spirit and scope of the invention, which is defined in the appended claims. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (5)
1. An intelligent farmland heavy metal online detection correction model management system, which is characterized by comprising:
the model registration module is used for providing registration service for correction models of various heavy metal detectors;
the spectrum input module is used for providing input interface services of original spectrum data of the soil standard sample and original spectrum data of the soil detection sample; the original spectrum data of the soil standard sample and the original spectrum data of the soil detection sample which need to be corrected all carry the equipment type number of the instrument and are used for being uniquely matched with a correction model;
the index selection module is used for selecting index sub-services of the correction model and deriving different correction models through different index sub-service combinations;
the data correction module is used for carrying out data correction calculation on the original spectrum;
the correction evaluation module is used for evaluating the correction effect of the standard sample after the original spectrum correction, and a model with the optimal evaluation result is used as a correction model corresponding to the heavy metal detector;
the spectrum output module is used for providing a corrected spectrum data output service; the detection sample correction spectrum data are directly stored into a correction spectrum database and used for subsequent analysis;
wherein the model registration module comprises:
the description specification unit is used for uniformly describing model indexes of correction models of various heavy metal detectors by adopting a structural modeling language and establishing a model description specification;
the standardized registration unit is used for realizing the full standardized registration of the existing correction model by identifying the standardized description file of the description specification unit;
the storage unit is used for uniformly storing the correction models registered by the standardized registration unit in a correction model database and storing the correction models according to a variety and identity sub-table of the correction models;
the index selection module comprises:
the standard sample selection unit is used for when a certain heavy metal detector does not have a corresponding correction model; the index selection module uses the input standard sample original spectrum data to traverse different methods corresponding to each standard correction model preset in the system, and selects corresponding required index sub-services from different methods, wherein the same index has different parameters and parameter values for selection;
the detection sample selection unit is used for matching the correction model in the correction model library according to the equipment number in the input detection sample original spectrum data when a certain heavy metal detector has a corresponding correction model, and then selecting model indexes and parameter values according to the index sub-service appointed during the registration of the correction model; if other experiments prove that the model index value has better choice, the standardized registration unit adjusts the original correction model index and the index value thereof, and uses the new correction model to correct.
2. The intelligent farmland heavy metal online detection and correction model management system according to claim 1, wherein the spectrum input module comprises:
the forming stage unit is used for calibrating each parameter and parameter value of the model by using the original spectrum data of the soil standard sample in the correcting model forming stage, and the calibrated correcting model can be stored in a correcting model database;
and the application stage unit is used for using the original spectrum data of the soil detection sample in a correction model application stage and calling a correction model in the correction model database to finish correction of the original spectrum information.
3. The intelligent farmland heavy metal online detection and correction model management system according to claim 1, wherein the data correction module comprises:
the first data correction unit is used for correcting and calculating the original spectrum data according to the index sub-service combination selected by the index selection module and outputting the corrected data to the correction evaluation module;
the second data correction unit is used for correcting and calculating the original spectrum data according to the index sub-service combination selected by the index selection module and outputting the corrected data to the spectrum output service.
4. The intelligent farmland heavy metal online detection and correction model management system according to claim 1, wherein the correction and evaluation module comprises:
the evaluation threshold unit is used for setting comprehensive correction evaluation indexes and thresholds, measuring correction evaluation results and thresholds, and selecting correction model results obtained by system automation evaluation if threshold conditions are met; if the evaluation result does not meet the threshold condition, repeating the index selection, the data correction and the correction evaluation until the evaluation result reaches the threshold condition;
and the comprehensive evaluation unit evaluates the correction result of each service combination of the spectrum of the standard sample by utilizing the comprehensive evaluation index and the threshold set by the evaluation threshold unit, and the correction model of the optimal evaluation result is used as the correction model of the heavy metal detection equipment and is registered by using the model registration module.
5. The intelligent farmland heavy metal online detection and correction model management system according to claim 1, wherein the spectrum output module comprises:
a spectrum output unit for outputting corrected spectrum data;
and a storage unit storing the corrected spectrum data for subsequent analysis.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110892464.8A CN113588572B (en) | 2021-08-04 | 2021-08-04 | Intelligent management system for farmland heavy metal online detection correction model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110892464.8A CN113588572B (en) | 2021-08-04 | 2021-08-04 | Intelligent management system for farmland heavy metal online detection correction model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113588572A CN113588572A (en) | 2021-11-02 |
| CN113588572B true CN113588572B (en) | 2024-03-19 |
Family
ID=78255058
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110892464.8A Active CN113588572B (en) | 2021-08-04 | 2021-08-04 | Intelligent management system for farmland heavy metal online detection correction model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113588572B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114486764B (en) * | 2022-01-26 | 2023-06-06 | 安徽新宇环保科技股份有限公司 | Agricultural non-point source pollution monitoring system based on full spectrum water quality analyzer |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101210875A (en) * | 2007-12-25 | 2008-07-02 | 浙江大学 | Damage-free measurement method for soil nutrient content based on near infrared spectra technology |
| CN101520421A (en) * | 2009-04-08 | 2009-09-02 | 北京农产品质量检测与农田环境监测技术研究中心 | Modeling method of soil heavy metal content detection model and application thereof |
| CN101738383A (en) * | 2008-11-07 | 2010-06-16 | 中国农业科学院农业环境与可持续发展研究所 | Mid-infrared spectrum-based method for quickly detecting contents of heavy metal elements in soil |
| CN101865828A (en) * | 2010-05-31 | 2010-10-20 | 湖南大学 | Method for maintaining predication capability of spectrum correction model of complex system |
| CN104897607A (en) * | 2015-06-18 | 2015-09-09 | 北京工商大学 | Food modeling and rapid detecting integration method and system adopting portable NIRS (near infrared spectroscopy) |
| CN108830253A (en) * | 2018-06-27 | 2018-11-16 | 中石化西南石油工程有限公司地质录井分公司 | Screening model method for building up, spectrum screening plant and method |
| CN108896492A (en) * | 2018-08-07 | 2018-11-27 | 华南农业大学 | Training method, storage medium and the terminal of PSO-BP neural network model |
| CN110018294A (en) * | 2019-05-14 | 2019-07-16 | 北京农业信息技术研究中心 | Heavy metal-polluted soil detects value correcting method, device and computer storage medium |
| CN111257228A (en) * | 2020-01-21 | 2020-06-09 | 黑龙江珍宝岛药业股份有限公司 | Hyperspectrum-based multi-index detection system for traditional Chinese medicine injection |
| CN111415715A (en) * | 2020-04-17 | 2020-07-14 | 北京北分瑞利分析仪器(集团)有限责任公司 | Intelligent correction method, system and device based on multivariate spectral data |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| RU2308684C1 (en) * | 2006-06-20 | 2007-10-20 | Общество с ограниченной ответственностью "ВИНТЕЛ" | Method of producing multi-dimension calibrating models |
-
2021
- 2021-08-04 CN CN202110892464.8A patent/CN113588572B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101210875A (en) * | 2007-12-25 | 2008-07-02 | 浙江大学 | Damage-free measurement method for soil nutrient content based on near infrared spectra technology |
| CN101738383A (en) * | 2008-11-07 | 2010-06-16 | 中国农业科学院农业环境与可持续发展研究所 | Mid-infrared spectrum-based method for quickly detecting contents of heavy metal elements in soil |
| CN101520421A (en) * | 2009-04-08 | 2009-09-02 | 北京农产品质量检测与农田环境监测技术研究中心 | Modeling method of soil heavy metal content detection model and application thereof |
| CN101865828A (en) * | 2010-05-31 | 2010-10-20 | 湖南大学 | Method for maintaining predication capability of spectrum correction model of complex system |
| CN104897607A (en) * | 2015-06-18 | 2015-09-09 | 北京工商大学 | Food modeling and rapid detecting integration method and system adopting portable NIRS (near infrared spectroscopy) |
| CN108830253A (en) * | 2018-06-27 | 2018-11-16 | 中石化西南石油工程有限公司地质录井分公司 | Screening model method for building up, spectrum screening plant and method |
| CN108896492A (en) * | 2018-08-07 | 2018-11-27 | 华南农业大学 | Training method, storage medium and the terminal of PSO-BP neural network model |
| CN110018294A (en) * | 2019-05-14 | 2019-07-16 | 北京农业信息技术研究中心 | Heavy metal-polluted soil detects value correcting method, device and computer storage medium |
| CN111257228A (en) * | 2020-01-21 | 2020-06-09 | 黑龙江珍宝岛药业股份有限公司 | Hyperspectrum-based multi-index detection system for traditional Chinese medicine injection |
| CN111415715A (en) * | 2020-04-17 | 2020-07-14 | 北京北分瑞利分析仪器(集团)有限责任公司 | Intelligent correction method, system and device based on multivariate spectral data |
Non-Patent Citations (1)
| Title |
|---|
| 华南地区土壤全钾含量高光谱反演模型研究;彭一平;《西南农业学报》;第32卷(第10期);2383-2389 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113588572A (en) | 2021-11-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Shirley et al. | HELP: the herschel extragalactic legacy project | |
| Lee et al. | The segue stellar parameter pipeline. i. description and comparison of individual methods | |
| CN104897607B (en) | Portable near infrared spectrum food modeling and quick detection integral method and system | |
| Desprez et al. | Euclid preparation-X. The Euclid photometric-redshift challenge | |
| CN105486655B (en) | The soil organism rapid detection method of model is intelligently identified based on infrared spectroscopy | |
| CN103235095B (en) | Water-injected meat detection method and device | |
| Fahrion et al. | The Fornax 3D project: Non-linear colour–metallicity relation of globular clusters | |
| CN102590129B (en) | Method for detecting content of amino acid in peanuts by near infrared method | |
| CN113588572B (en) | Intelligent management system for farmland heavy metal online detection correction model | |
| CN109580534A (en) | Substance classes recognition methods, terahertz time-domain spectroscopy instrument system and terminal device | |
| Worley et al. | The AMBRE Project: Stellar parameterisation of the ESO: UVES archived spectra | |
| CN106950192A (en) | A kind of method of Contents of Main Components quick detection in vegetable protein beverage based on near-infrared spectral analysis technology | |
| CN106018331B (en) | The method for estimating stability and pretreatment optimization method of multi-channel spectral system | |
| CN109992861A (en) | A kind of near infrared spectrum modeling method | |
| CN110503156A (en) | A kind of multivariate calibration characteristic wavelength selection method based on least correlativing coefficient | |
| CN103018178A (en) | LPP (Local Preserving Projection)-based Infrared spectrometer calibration method | |
| CN106126879B (en) | A kind of soil near-infrared spectrum analysis prediction technique based on rarefaction representation technology | |
| CN116818687B (en) | Soil organic carbon spectrum prediction method and device based on spectrum guide integrated learning | |
| CN109670531A (en) | A kind of denoising method of the near infrared light spectrum signal based on Hodrick-Prescott filter | |
| Sandberg et al. | Trident: A three-pronged galaxy survey-I. Lyman alpha emitting galaxies at z~ 2 in GOODS North | |
| CN108489928B (en) | Method for detecting textile fiber components by short-wave infrared extinction spectrum | |
| CN105067558B (en) | Near-infrared Qualitive test feature extracting method | |
| CN109596545B (en) | Quality control method of full-spectrum water quality monitoring equipment | |
| CN110231306A (en) | A kind of method of lossless, the quick odd sub- seed protein content of measurement | |
| Tanaka et al. | Star formation activities of galaxies in the large-scale structures at z= 1.2 |
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 |