Environmental monitoring equipment calibration method and system
Through the collection and analysis of environmental samples and historical information, the environmental monitoring equipment is calibrated using the content prediction model, which solves the detection inaccurate problem caused by spectral intervention in the natural environment, and improves the accuracy of the equipment and resource utilization efficiency.
Patent Information
- Application Number
- CN202410426637.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-04-10
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2044-04-10
AI Technical Summary
Existing environmental monitoring equipment is inaccurately detected under spectral intervention in the natural environment, making it difficult to make effective corrections.
By collecting environmental samples and historical content information, the content prediction model is used to determine whether the device needs calibration, and train it based on the comparison content information and historical information, and adjust the duration of the detection cycle to adapt to environmental changes.
It improves the accuracy and reliability of environmental monitoring data, promptly corrects equipment deviations, reasonably arranges resource use, reduces waste of computing resources, and properly disposes samples to reduce environmental pollution.
Smart Images

Figure CN118329802B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of environmental monitoring equipment, and in particular to an environmental monitoring equipment calibration method and system. Background Art
[0002] In the related technology, CN114486765A discloses a collaborative correction method and system for a full-spectrum water quality analyzer, which belongs to the field of data processing technology of full-spectrum water quality analyzer, and includes a server, an integration module, a water quality detection module and a collaborative correction module; the integration module is used to set up a collaborative correction device; the water quality detection module is used to perform water quality detection through the full-spectrum water quality analyzer, obtain a water quality detection plan, identify the location of the water quality detection point in the water quality detection plan, set the full-spectrum water quality analyzer at the water quality detection point, perform water quality detection, obtain detection data, and name the full-spectrum water quality analyzers located on the same single tributary according to the corresponding collaborative correction device; realize real-time detection and correction of water quality through the full-spectrum water quality analyzer, avoid the inconvenience and high cost of sending the full-spectrum water quality analyzer to the laboratory for calibration, and realize accurate measurement of water quality in seconds.
[0003] CN102121902A discloses an online Raman spectrometer calibration device and a calibration method thereof. The laser power of a laser of an online Raman detection system is detected as a reference power, and the position of a Raman spectrum peak obtained by a neon monochromatic light source is used as a reference position. The laser power and spectrum position of the online Raman spectrometer are calibrated in real time, so that when the online Raman spectrum works for a long time, the correction when the laser power decreases or deviates and the correction when the spectrum level deviates are achieved, thereby achieving long-term effective operation of the online Raman spectrometer in harsh environments, and ensuring that the online Raman spectrometer can effectively perform sample detection on site in the long term.
[0004] Therefore, in the related art, the content of various substances in the environment can be detected by a spectrometer. However, since the spectrometer may be affected by the ambient light of the natural environment in which it is located, there are some detection errors, resulting in inaccurate detection, and the correction method in the related art is difficult to calibrate the spectrometer in various natural environments.
[0005] The information disclosed in the background technology section of this application is only intended to deepen the understanding of the general background technology of this application, and should not be regarded as an admission or any form of suggestion that the information constitutes the prior art already known to those skilled in the art. Summary of the invention
[0006] The present invention provides an environmental monitoring equipment calibration method and system, which can solve the technical problem of inaccurate detection of environmental monitoring equipment.
[0007] According to a first aspect of the present invention, there is provided a method for calibrating an environmental monitoring device, comprising:
[0008] At the beginning of the current detection cycle, environmental samples are collected and saved;
[0009] Obtain historical content information of multiple substances in environmental samples at the beginning of the previous detection cycle;
[0010] Using an environmental monitoring spectral device, testing environmental samples at the beginning of a previous testing cycle to obtain comparative spectral data, and inputting the comparative spectral data into a content prediction model to obtain comparative content information of multiple substances;
[0011] Determining whether the environmental monitoring spectroscopy device needs to be calibrated in a current detection cycle according to the comparative content information and the historical content information;
[0012] If the environmental monitoring spectral device needs to be calibrated in the current detection cycle, the content prediction model is trained according to the comparative content information, the historical content information and the preset calibration sample to obtain the trained content prediction model and the calibrated environmental monitoring spectral device;
[0013] Determine the duration of the current detection cycle according to the comparison content information, the historical content information, and the environmental sample at the start time of the current detection cycle;
[0014] Performing environmental monitoring during the current detection cycle according to the calibrated environmental monitoring spectroscopy device, the trained content prediction model, and the duration of the current detection cycle;
[0015] The environmental sample at the start time of the previous detection cycle is moved to the recovery area and recovered.
[0016] According to a second aspect of the present invention, there is provided an environmental monitoring equipment calibration system, the system comprising:
[0017] The environmental sample module is used to collect and save environmental samples at the beginning of the current detection cycle;
[0018] A historical content information module is used to obtain historical content information of multiple substances in environmental samples at the beginning of the previous detection cycle;
[0019] A comparative content information module is used to detect the environmental sample at the start time of the previous detection cycle through an environmental monitoring spectrum device to obtain comparative spectrum data, and input the comparative spectrum data into a content prediction model to obtain comparative content information of multiple substances;
[0020] A calibration module is used to determine whether the environmental monitoring spectroscopy device needs to be calibrated in the current detection cycle according to the comparison content information and the historical content information;
[0021] A prediction model and spectral device module, for training the content prediction model according to the comparative content information, the historical content information and a preset calibration sample to obtain a trained content prediction model and a calibrated environmental monitoring spectral device if the environmental monitoring spectral device needs to be calibrated in the current detection cycle;
[0022] A duration module, used to determine the duration of the current detection cycle according to the comparison content information, the historical content information, and the environmental sample at the start time of the current detection cycle;
[0023] An environmental monitoring module, configured to perform environmental monitoring within the current detection cycle according to the calibrated environmental monitoring spectral device, the trained content prediction model and the duration of the current detection cycle;
[0024] The recycling module is used to move the environmental sample at the start time of the previous detection cycle to the waiting recycling area and recycle it.
[0025] According to a third aspect of the present invention, there is provided an environmental monitoring device calibration device, comprising: a processor; a memory for storing processor executable instructions; wherein the processor is configured to call the instructions stored in the memory to execute the environmental monitoring device calibration method.
[0026] According to a fourth aspect of the present invention, there is provided a computer-readable storage medium having computer program instructions stored thereon, wherein the computer program instructions implement the environmental monitoring device calibration method when executed by a processor.
[0027] Technical effect: According to the present invention, by comparing the comparative content information obtained by testing the environmental samples of the previous detection cycle and the historical content information detected in the previous detection cycle, it is judged whether the natural environment in which the environmental monitoring spectrum device is located has changed within a detection cycle. If the environmental monitoring spectrum device is no longer adapted to the natural environment, the content prediction model can be trained through the calibration sample to improve the accuracy of the content prediction model, thereby improving the accuracy and reliability of the environmental monitoring data. The calibrated environmental monitoring spectrum device can improve the accuracy during the working process and correct the deviation in time. By judging whether the environmental monitoring spectrum device is adapted to the natural environment and determining the duration of the detection cycle, the use of resources and equipment can be reasonably arranged to reduce the waste of computing resources. Recycling can properly dispose of samples and reduce environmental pollution. When judging whether the environmental monitoring spectrum device needs to be calibrated, the substance content information in multiple sets based on the comparative content information and the historical content information and the reasons for the existence of errors are comprehensively considered, and the content error of each substance is determined based on the sensitivity error, so as to more comprehensively evaluate the accuracy and stability of the environmental monitoring spectrum device. The weighted error is a quantitative indicator to measure the difference between the historical actual content and the comparative content. By determining the weighted error, the error degree of the environmental monitoring spectral equipment in the current detection cycle can be objectively reflected, so as to judge whether the environmental monitoring spectral equipment is adapted to the natural environment of the current detection cycle, and then accurately judge whether the environmental monitoring spectral equipment needs to be calibrated in the current detection cycle. When training the content prediction model, the measurement error of the calibration sample can be weighted by the time period of natural environment change corresponding to the measurement error of the calibration sample in the process of determining the loss function, and the error between the historical content information and the comparative content information can be weighted by the time period of natural environment change corresponding to the historical content information, so as to accurately determine the cause of the error, and reduce the error through training, so as to improve the accuracy of the content prediction model. When determining the duration of the current detection cycle, the duration of the current detection cycle can be adjusted according to the environmental impact coefficient and the material content change coefficient to reflect the overall changes in environmental changes and material content changes. If the overall changes are large, the duration of the detection cycle can be shortened to detect the changing environment at a higher frequency, thereby improving the detection accuracy and real-time performance. On the contrary, if the overall changes are small, the duration of the detection cycle is increased to save computing power. Adjust the duration of the detection cycle according to actual conditions to make the detection more adaptable, flexible and economical, and improve the accuracy and efficiency of the detection.
[0028] It should be understood that the above general description and the following detailed description are exemplary and explanatory only and do not limit the present invention. Other features and aspects of the present invention will become more apparent from the following detailed description of exemplary embodiments with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS
[0029] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other embodiments can be obtained based on these drawings without creative work.
[0030] Figure 1 A schematic diagram exemplarily shows a flow chart of a method for calibrating an environmental monitoring device according to an embodiment of the present invention;
[0031] Figure 2 A block diagram of an environmental monitoring device calibration system according to an embodiment of the present invention is exemplarily shown. DETAILED DESCRIPTION
[0032] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0033] The technical solution of the present invention is described in detail with specific embodiments below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.
[0034] Figure 1 A flow chart of a method for calibrating an environmental monitoring device according to an embodiment of the present invention is exemplarily shown, the method comprising:
[0035] Step S101, at the beginning of the current detection cycle, collect and save environmental samples;
[0036] Step S102, obtaining historical content information of multiple substances in the environmental sample at the start time of the previous detection cycle;
[0037] Step S103, using an environmental monitoring spectrum device to detect the environmental sample at the start time of the previous detection cycle, obtain comparative spectrum data, and input the comparative spectrum data into a content prediction model to obtain comparative content information of multiple substances;
[0038] Step S104, determining whether the environmental monitoring spectroscopy device needs to be calibrated in the current detection cycle according to the comparison content information and the historical content information;
[0039] Step S105, if the environmental monitoring spectral device needs to be calibrated in the current detection cycle, the content prediction model is trained according to the comparative content information, the historical content information and the preset calibration sample to obtain the trained content prediction model and the calibrated environmental monitoring spectral device;
[0040] Step S106, determining the duration of the current detection cycle according to the comparison content information, the historical content information, and the environmental sample at the start time of the current detection cycle;
[0041] Step S107, performing environmental monitoring in the current detection cycle according to the calibrated environmental monitoring spectral equipment, the trained content prediction model and the duration of the current detection cycle;
[0042] Step S108, moving the environmental sample at the start time of the previous detection cycle to the waiting recycling area and recycling it.
[0043] According to the environmental monitoring equipment calibration method of an embodiment of the present invention, by comparing the comparative content information obtained by testing the environmental samples of the previous detection cycle and the historical content information detected in the previous detection cycle, it is determined whether the natural environment in which the environmental monitoring spectral equipment is located has changed within a detection cycle. If the environmental monitoring spectral equipment is no longer adapted to the natural environment, the content prediction model can be trained through the calibration sample to improve the accuracy of the content prediction model, thereby improving the accuracy and reliability of the environmental monitoring data. The calibrated environmental monitoring spectral equipment can improve the accuracy during the working process and correct the deviation in time. By judging whether the environmental monitoring spectral equipment is adapted to the natural environment and determining the duration of the detection cycle, the use of resources and equipment can be reasonably arranged to reduce the waste of computing resources. Recycling can properly dispose of samples and reduce environmental pollution.
[0044] According to one embodiment of the present invention, in step S101, at the beginning of each detection cycle, samples in the environment are collected and saved for subsequent use. The collected environmental samples may include samples of air, water, soil or other related substances, depending on the specific environmental monitoring requirements.
[0045] According to one embodiment of the present invention, in step S102, during the environmental monitoring process, an environmental sample at the start of the previous cycle has been collected and saved. The environmental sample contains historical content information of various substances, such as the concentration and content of various chemicals, pollutants or other substances. The historical content information provides an important reference for subsequent environmental monitoring, calibration and prediction.
[0046] According to one embodiment of the present invention, in step S103, the environmental monitoring spectral device is a device for analyzing the spectral characteristics of different substances in an environmental sample, and the content of multiple substances in the environment can be determined by spectral analysis. At the beginning of the current detection cycle, the device obtains comparative spectral data by analyzing the spectral characteristics of the environmental sample at the beginning of the previous detection cycle. The content prediction model is a model for predicting the content of multiple substances in an environmental sample. By inputting the comparative spectral data into the model, the content information of multiple substances in the environmental sample at the beginning of the previous detection cycle, that is, comparative content information, is obtained.
[0047] According to one embodiment of the present invention, in step S104, the comparison content information and the historical content information are the content information of multiple substances in the same environmental sample. If the difference between the comparison content information and the historical content information is relatively obvious, it means that the natural environment in which the environmental monitoring spectrum device is located has undergone a significant change in a detection cycle, resulting in a significant difference in the content information measured by the environmental monitoring spectrum device at the beginning of the current detection cycle and the beginning of the previous detection cycle. In other words, due to changes in the natural environment (for example, the lighting environment), errors occur in the environmental monitoring spectrum device. The historical content information is the content information measured by the environmental monitoring spectrum device when adapting to the natural environment. If the error between the comparison content information and the historical content information is large, the environmental monitoring spectrum device cannot adapt to the natural environment of the current detection cycle and needs to be calibrated to improve the accuracy and reliability of the measurement results. If the error between the comparison content information and the historical content information is small, the environmental monitoring spectrum device does not need to be calibrated to save computing power.
[0048] According to one embodiment of the present invention, step S104 includes: determining, based on the comparative content information and the historical content information, a first type set of multiple substances corresponding to the historical content information, and a second type set of multiple substances corresponding to the comparative content information; determining the intersection of the first type set and the second type set; determining a third type set included in the first type set but not included in the intersection; determining a fourth type set included in the second type set but not included in the intersection; and determining whether the environmental monitoring spectral equipment needs to be calibrated in the current detection cycle based on the comparative content information, the historical content information, the intersection, the third type set and the fourth type set.
[0049] According to one embodiment of the present invention, different sets of multiple substances in the environment are distinguished based on comparative content information and historical content information. The intersection of the first type set and the second type set represents a set of substances that are determined to exist in both the historical content information and the comparative content information. The third type set represents a set of substances that exist in the historical content information but do not appear in the comparative content information. The fourth type set represents a set of substances that appear in the comparative content information but do not appear in the historical content information. By comparing the substance contents and sets of different data sources, it is determined whether there are substances or changes that need to be calibrated in the current detection cycle, thereby determining whether equipment calibration is required.
[0050] According to an embodiment of the present invention, determining whether the environmental monitoring spectrum device needs to be calibrated in the current detection cycle according to the comparison content information, the historical content information, the intersection, the third type set and the fourth type set includes: obtaining a weighted error E of the environmental monitoring spectrum device in two adjacent detection cycles according to formula (1),
[0051]
[0052] Among them, C 1,i is the historical content information of the i-th type of substance in the intersection, C 2,i is the comparative content information of the i-th type of substance in the intersection, I is the total number of types in the intersection, C 3,j is the historical content information of the jth type of substance in the third type set, C ET is the sensitivity error, n 3 is the total number of types in the third type set, C 4,k is the content information of the kth type of substance in the fourth type set, n 4 is the total number of types in the fourth type set, if is the conditional function, i≤I, j≤n 3 , k≤n 4 , and i, j, k, I, n 3 and n 4 are all positive integers; when the weighted error is greater than or equal to the preset error threshold, it is determined that the environmental monitoring spectrum device needs to be calibrated in the current detection cycle.
[0053] According to one embodiment of the present invention, in formula (1), It is the absolute value of the difference between the historical content information and the comparative content information of each type of substance in the intersection, and they are accumulated and summed. Indicates that when the historical content information of the j-th type of substance in the third type set is greater than the sensitivity error, the conditional function value is the historical content information of the j-th type of substance in the third type set and 1-C ET, otherwise, the conditional function value is the historical content information of the j-th type of substance in the third type set. If the historical content information of the j-th type of substance in the third type set is greater than the sensitivity error, it means that the environmental monitoring spectral device did not detect the j-th type of substance in the third type set due to the detection error in the current detection cycle. Otherwise, it may be due to the sensitivity error that the j-th type of substance in the third type set was not detected. If the historical content information of the j-th type of substance in the third type set is greater than the sensitivity error, then It means that the historical content information is adjusted to eliminate the interference of sensitivity error, so as to obtain the true error of the content of the j-th type of substance in the third type set and improve the accuracy of the calculation. Otherwise, the historical content information of the j-th type of substance in the third type set that is not detected by the environmental monitoring spectral equipment can be used as the true error. The actual error of each type of substance in the third type set is accumulated and summed. Since the historical content information can be considered as accurate content information, the types of substances in the fourth type set can be considered as substances detected due to detection errors in the environmental monitoring spectral equipment. Therefore, After the content information of each type of substance in the fourth type set is processed to eliminate the interference of sensitivity errors, the sum is performed. That is, the content information of each type of substance in the fourth type set is the content information of the substance detected by mistake due to the environmental monitoring spectrum device not adapting to the environment. Therefore, the content information of each type of substance in the fourth type set can be processed to eliminate the interference of sensitivity errors to sum the real errors. The above three items are added together to obtain the weighted error of the environmental monitoring spectrum device in two adjacent detection cycles. If the weighted error is greater than or equal to the preset error threshold, it is determined that the environmental monitoring spectrum device needs to be calibrated in the current detection cycle.
[0054] In this way, the substance content information in multiple sets based on comparative content information and historical content information and the reasons for the existence of errors are comprehensively considered, and the content error of each substance is determined based on the sensitivity error, so as to more comprehensively evaluate the accuracy and stability of the environmental monitoring spectral equipment. The weighted error is a quantitative indicator to measure the difference between the historical actual content and the comparative content. By determining the weighted error, the error degree of the environmental monitoring spectral equipment in the current detection cycle can be objectively reflected, so as to judge whether the environmental monitoring spectral equipment is adapted to the natural environment of the current detection cycle, and then accurately judge whether the environmental monitoring spectral equipment needs to be calibrated in the current detection cycle.
[0055] According to one embodiment of the present invention, in step S105, when training the content prediction model, the difference between the measurement result of the environmental monitoring spectral device and the actual content can be compared based on the preset calibration sample, so as to adjust and optimize the parameters of the model. Among them, the preset calibration sample is accurately measured, certified or standardized, and the accurate content of its various substances is known. After training, a trained content prediction model will be obtained, which can more accurately detect the content information of various substances in the environmental sample. At the same time, the calibrated environmental monitoring spectral device is also updated, and can provide accurate and reliable results during measurement.
[0056] According to one embodiment of the present invention, step S105 includes: detecting the sample spectral data of the preset calibration sample through the environmental monitoring spectral equipment; inputting the sample spectral data into the content prediction model to obtain the sample content information of multiple substances in the preset calibration sample; determining the loss function of the content prediction model according to the sample content information, the annotation information of the preset calibration sample, the comparative content information and the historical content information; training the content prediction model according to the loss function to obtain the trained content prediction model; obtaining the calibrated environmental monitoring spectral equipment according to the trained content prediction model.
[0057] According to one embodiment of the present invention, the calibration sample is a sample with known substance content, which is used to train and verify the content prediction model. The loss function is used to feedback and adjust the model parameters, thereby optimizing the accuracy of the model. According to the trained content prediction model, the environmental monitoring spectroscopy equipment is calibrated, and the calibrated equipment can more accurately detect the content of the substance in the sample, thereby improving the accuracy and reliability of the equipment.
[0058] According to an embodiment of the present invention, determining the loss function of the content prediction model according to the sample content information, the annotation information of the preset calibration sample, the comparative content information and the historical content information includes: determining the loss function Loss of the content prediction model according to formula (2):
[0059]
[0060] Among them, Ω 1 is a first type set, Ω 2 is the second type set, Ω 1 ∪Ω 2 is the union of the first type set and the second type set, count(Ω 1 ∪Ω 2 ) is the total number of types in the union of the first type set and the second type set, C 1,s is the historical content information of the s-th type of substance in the union of the first type set and the second type set, C2,s is the comparative content information of the s-th type of substance in the union of the first type set and the second type set, ΔT is the duration of the previous detection cycle, T T is the duration of the preset monitoring period, C ET is the sensitivity error, Ω 5 is a fifth type set of multiple substances contained in the preset calibration sample determined according to the annotation information, Ω 6 is a sixth type set of multiple substances contained in the preset calibration sample determined according to the sample content information, Ω 5 ∪Ω 6 is the union of the fifth type set and the sixth type set, count(Ω 5 ∪Ω 6 ) is the total number of types in the union of the fifth type set and the sixth type set, C b,t is the content information of the substance of the tth type in the union of the fifth type set and the sixth type set determined according to the annotation information, C y,t is the sample content information of the t-th type of substance in the union of the fifth type set and the sixth type set, ΔT L is the time interval between the last training of the content prediction model and the start of the current detection cycle, s≤count(Ω 1 ∪Ω 2 ), t≤count(Ω 5 ∪Ω 6 ), and s and t are both positive integers.
[0061] According to one embodiment of the present invention, in formula (2), T T The duration of the preset monitoring time period can be set to 1 day, one week, etc., and the present invention is not limited to this. The ratio of the absolute value of the difference between the content information of the t-th type of substance in the union of the fifth type set and the sixth type set determined according to the annotation information and the sample content information of the t-th type of substance in the union of the fifth type set and the sixth type set minus the sensitivity error to 1. This ratio can eliminate the influence of the sensitivity error and amplify the content information difference, thereby improving the training efficiency. is the ratio of the time interval between the last training of the content prediction model and the start of the current detection cycle to the duration of the preset monitoring time period, which is used to reasonably weight the error of the calibration time interval in the loss function. That is, it can be considered that after the last training of the content prediction model, the environmental monitoring spectroscopy equipment can adapt to the natural environment during the last training. Therefore, the accuracy of the content information of various substances in the calibration sample measured by the environmental monitoring spectroscopy equipment after the last calibration is higher, and the above error This is because at ΔTL The error is caused by the change of natural environment within the time period, which causes the environmental monitoring spectrum equipment to be unable to adapt to the current natural environment. Therefore, the error can be weighted by the time period corresponding to it.
[0062] According to one embodiment of the present invention, in formula (2), It is the ratio of the absolute value of the difference between the historical content information of the s-th type of substance in the union of the first type set and the second type set and the comparative content information of the s-th type of substance in the union of the first type set and the second type set and 1 minus the sensitivity error. This ratio can eliminate the influence of the sensitivity error and amplify the content information difference, thereby improving the training efficiency. It is the ratio of the duration of the previous detection cycle to the duration of the preset monitoring time period, which is used to reasonably weight the error of the detection time interval in the loss function. In the previous detection cycle, the environmental monitoring spectral equipment was able to adapt to the natural environment, so the historical content information measured was accurate. This is caused by the change of the natural environment within a detection cycle, which causes the environmental monitoring spectrum equipment to be unsuitable for the current natural environment. Therefore, the error can be weighted by the time period corresponding to it.
[0063] According to an embodiment of the present invention, if the previous detection cycle also performed a calibration process on the environmental monitoring spectrum device, then ΔT L =ΔT, otherwise, ΔT L >ΔT. The above two terms can be added together to obtain the loss function of the content prediction model.
[0064] According to one embodiment of the present invention, during the process of training the content prediction model, the loss function can be back-propagated and some parameters within the model can be adjusted to reduce the value of the loss function, thereby improving the accuracy and reliability of the content prediction model, obtaining a trained content prediction model, and further obtaining a calibrated environmental monitoring spectral device.
[0065] In this way, in the process of determining the loss function, the measurement error of the calibration sample can be weighted by the time period of natural environment changes corresponding to the measurement error of the calibration sample, and the error between the historical content information and the comparative content information can be weighted by the time period of natural environment changes corresponding to the historical content information, so as to accurately determine the cause of the error, and reduce the error in a targeted manner through training, thereby improving the accuracy of the content prediction model.
[0066] According to one embodiment of the present invention, in step S106, based on the comparative content information and historical content information of the environmental sample at the start of the previous detection cycle, it can be understood whether the environmental monitoring equipment adapts to environmental changes. The environmental monitoring equipment that adapts to environmental changes detects the environmental sample at the start of the current detection cycle, and can observe the actual content of the substance in the environment, thereby determining the duration of the current detection cycle. That is, if it is found that the content of the substance in the environment changes greatly or fluctuates more frequently compared to the previous detection cycle, it may be necessary to shorten the detection cycle in order to monitor the changes in the environment more timely. On the contrary, if the content of the substance in the environment changes little or is stable, a longer detection cycle can be selected.
[0067] According to one embodiment of the present invention, step S106 includes: obtaining the measured content information of multiple substances in the environmental sample at the start time of the current detection cycle through a calibrated environmental monitoring spectral device; determining the substance content variation coefficient based on the measured content information and the historical content information; determining the environmental impact coefficient based on the comparative content information and the historical content information; determining the duration of the current detection cycle based on the substance content variation coefficient and the environmental impact coefficient.
[0068] According to one embodiment of the present invention, the measured content information and the historical content information are measured by a spectral device adapted to the environment, that is, both content information are accurate. The coefficient of variation of the material content thus obtained can describe the trend of the change of the material content between different time points, so as to understand the change of the material in the environment. The comparative content information is measured by the spectral device before calibration, so the difference between the comparative content information and the historical content information represents the impact of environmental factors on the spectral device. The environmental impact coefficient thus obtained can measure the change of the material content measured by the spectral device due to environmental factors, and then reflect the change of the environment. By considering the change trend of the material content and the influence of environmental factors on the duration of the previous detection cycle, the duration of the current detection cycle can be determined. If the change of the material content or the change of the environment is large, the detection cycle can be appropriately shortened, so that the environment with faster changes can be detected more frequently. Otherwise, the detection cycle can be appropriately increased, so as to reduce the number of detections of the environment with slower changes and reduce the computing power occupied.
[0069] According to an embodiment of the present invention, determining the duration of the current detection cycle according to the substance content variation coefficient and the environmental impact coefficient includes: determining the duration of the current detection cycle ΔT according to formula (3): N ,
[0070] ΔT N =ΔT-if(H×M≥ε, HMΔT, -HMΔT) (3)
[0071] in, is the environmental impact coefficient, is the coefficient of variation of the substance content, Ω 1 is a first type set, Ω 2 is the second type set, Ω 1 ∪Ω 2 is the union of the first type set and the second type set, count(Ω 1 ∪Ω 2 ) is the total number of types in the union of the first type set and the second type set, C 1,s is the historical content information of the s-th type of substance in the union of the first type set and the second type set, C 2,s is the comparative content information of the s-th type of substance in the union of the first type set and the second type set, ΔT is the duration of the previous detection cycle, Ω 7 is the seventh type set of various substances corresponding to the measured content information, Ω 1 ∪Ω 7 is the union of the first type set and the seventh type set, count(Ω 1 ∪Ω 7 ) is the total number of types in the union of the first type set and the seventh type set, is the historical content information of the h-th type of substance in the union of the first type set and the seventh type set, is the measured content information of the hth type of substance in the union of the first type set and the seventh type set, s≤count(Ω 1 ∪Ω 2 ), h≤count(Ω 1 ∪Ω 7 ), and s and h are both positive integers.
[0072] According to one embodiment of the present invention, if (C 1,s ≠0, C 1,s , C 2,s ) indicates that when the historical content information of the s-th type of substance in the union of the first type set and the second type set is not equal to 0, the conditional function value is the historical content information of the s-th type of substance in the union of the first type set and the second type set; otherwise, the conditional function value is the comparative content information of the s-th type of substance in the union of the first type set and the second type set. Relative difference between historical content information of the s-th type of substance in the union of the first type set and the second type set and comparative content information of the s-th type of substance in the union of the first type set and the second type set. This value is based on the difference between the historical content change and the comparative content, taking into account the influence of environmental factors on the change of substance content measured by the spectral equipment, and can further reflect the change of the environment, that is, the greater the change of the natural environment, the greater the above relative difference. It indicates the average value of the relative difference, that is, the environmental impact coefficient. The larger the value, the greater the impact of environmental changes on the test results of the substance content, which reflects that the environmental changes are relatively large.
[0073] According to one embodiment of the present invention, When the historical content information of the h-th type of substance in the union of the first type set and the seventh type set is not equal to 0, the conditional function value is the historical content information of the h-th type of substance in the union of the first type set and the seventh type set; otherwise, the conditional function value is the measured content information of the h-th type of substance in the union of the first type set and the seventh type set. Relative difference between historical content information of the h-th type of substance in the union of the first type set and the seventh type set and the measured content information of the h-th type of substance in the first type set and the union of the first type set and the seventh type set. This value can reflect the change of the content of the substance in the environment within a detection cycle according to the difference between the historical content change and the measured content. The larger the value, the greater the change in the substance content within a detection period.
[0074] According to one embodiment of the present invention, in formula (3), H×M is the product of the above two items, which can represent the overall change of the content of the substance and the environment. HMΔT is the overall change multiplied by the duration of the previous detection cycle, which can represent the change amplitude of the duration of the current detection cycle relative to the duration of the previous detection cycle under the influence of the above overall change. if(H×M≥ε, HMΔT, -HMΔT) means that when the total change of the content of the substance is greater than or equal to ε, the conditional function value is the change amplitude, otherwise, the conditional function value is the negative value of the change amplitude. ΔT-if(H×M≥ε, HMΔT, -HMΔT) means that if the total change of the content of the substance is greater than or equal to ε, then the duration of the current detection cycle will be reduced by the value of the change ratio, that is, in the case of a large overall change, the duration of the current detection cycle can be shortened relative to the duration of the previous detection cycle, and the shortening amplitude is the above change amplitude. If the total change in the substance content is less than ε, the duration of the current detection cycle will be added with the value of the change ratio (a negative value), that is, when the overall change is small, the duration of the current detection cycle can increase relative to the duration of the previous detection cycle, and the increase is the above-mentioned change range.
[0075] In this way, the duration of the current detection cycle can be adjusted according to the environmental impact coefficient and the material content change coefficient to reflect the overall changes in the environment and material content. If the overall changes are large, the duration of the detection cycle can be shortened to detect the changing environment at a higher frequency to improve the accuracy and real-time performance of the detection. Conversely, if the overall changes are small, the duration of the detection cycle is increased to save computing power. Adjusting the duration of the detection cycle according to the actual situation makes the detection more adaptable, flexible and economical, and improves the accuracy and efficiency of the detection.
[0076] According to one embodiment of the present invention, in step S107, after the content prediction model of the environmental monitoring spectral device is trained, a calibrated environmental monitoring spectral device can be obtained, which has been adjusted and optimized and can provide accurate and reliable results during the measurement process. According to the determined duration of the current detection cycle, environmental monitoring will be performed during this period, and the accuracy of the content information of the multiple substances detected at this time is higher.
[0077] According to one embodiment of the present invention, in step S108, during the environmental monitoring process, the environmental sample at the beginning of the previous detection cycle is usually retained for subsequent analysis and comparison. However, after using the sample for comparison in the current detection cycle, it can be moved to the recovery area. The recovery area is a designated area or container for temporarily storing samples that have completed environmental monitoring. Recovery usually includes appropriate disposal and treatment measures so that the various substances in the sample will not cause pollution or harm to the environment.
[0078] According to the environmental monitoring equipment calibration method of the embodiment of the present invention, by comparing the comparative content information obtained by detecting the environmental samples of the previous detection cycle and the historical content information detected in the previous detection cycle, it is judged whether the natural environment in which the environmental monitoring spectrum equipment is located has changed within a detection cycle. If the environmental monitoring spectrum equipment is no longer adapted to the natural environment, the content prediction model can be trained through the calibration sample to improve the accuracy of the content prediction model, thereby improving the accuracy and reliability of the environmental monitoring data. The calibrated environmental monitoring spectrum equipment can improve the accuracy during the working process and correct the deviation in time. By judging whether the environmental monitoring spectrum equipment is adapted to the natural environment and determining the duration of the detection cycle, the use of resources and equipment can be reasonably arranged to reduce the waste of computing resources. Recycling can properly dispose of samples and reduce environmental pollution. When judging whether the environmental monitoring spectrum equipment needs to be calibrated, the material content information in multiple sets based on the comparative content information and the historical content information and the reasons for the existence of errors are comprehensively considered, and the content error of each substance is determined based on the sensitivity error, so as to more comprehensively evaluate the accuracy and stability of the environmental monitoring spectrum equipment. The weighted error is a quantitative indicator to measure the difference between the historical actual content and the comparative content. By determining the weighted error, the error degree of the environmental monitoring spectral equipment in the current detection cycle can be objectively reflected, so as to judge whether the environmental monitoring spectral equipment is adapted to the natural environment of the current detection cycle, and then accurately judge whether the environmental monitoring spectral equipment needs to be calibrated in the current detection cycle. When training the content prediction model, the measurement error of the calibration sample can be weighted by the time period of natural environment change corresponding to the measurement error of the calibration sample in the process of determining the loss function, and the error between the historical content information and the comparative content information can be weighted by the time period of natural environment change corresponding to the historical content information, so as to accurately determine the cause of the error, and reduce the error through training, so as to improve the accuracy of the content prediction model. When determining the duration of the current detection cycle, the duration of the current detection cycle can be adjusted according to the environmental impact coefficient and the material content change coefficient to reflect the overall changes in environmental changes and material content changes. If the overall changes are large, the duration of the detection cycle can be shortened to detect the changing environment at a higher frequency, thereby improving the detection accuracy and real-time performance. On the contrary, if the overall changes are small, the duration of the detection cycle is increased to save computing power. Adjust the duration of the detection cycle according to actual conditions to make the detection more adaptable, flexible and economical, and improve the accuracy and efficiency of the detection.
[0079] Figure 2 A block diagram of an environmental monitoring device calibration system according to an embodiment of the present invention is exemplarily shown, the system comprising:
[0080] The environmental sample module is used to collect and save environmental samples at the beginning of the current detection cycle;
[0081] A historical content information module is used to obtain historical content information of multiple substances in environmental samples at the beginning of the previous detection cycle;
[0082] A comparative content information module is used to detect the environmental sample at the start time of the previous detection cycle through an environmental monitoring spectrum device to obtain comparative spectrum data, and input the comparative spectrum data into a content prediction model to obtain comparative content information of multiple substances;
[0083] A calibration module is used to determine whether the environmental monitoring spectroscopy device needs to be calibrated in the current detection cycle according to the comparison content information and the historical content information;
[0084] A prediction model and spectral device module, for training the content prediction model according to the comparative content information, the historical content information and a preset calibration sample to obtain a trained content prediction model and a calibrated environmental monitoring spectral device if the environmental monitoring spectral device needs to be calibrated in the current detection cycle;
[0085] A duration module, used to determine the duration of the current detection cycle according to the comparison content information, the historical content information, and the environmental sample at the start time of the current detection cycle;
[0086] An environmental monitoring module, configured to perform environmental monitoring within the current detection cycle according to the calibrated environmental monitoring spectral device, the trained content prediction model and the duration of the current detection cycle;
[0087] The recycling module is used to move the environmental sample at the start time of the previous detection cycle to the waiting recycling area and recycle it.
[0088] According to one embodiment of the present invention, there is provided an environmental monitoring device calibration device, comprising: a processor; a memory for storing processor executable instructions; wherein the processor is configured to call the instructions stored in the memory to execute the environmental monitoring device calibration method.
[0089] According to one embodiment of the present invention, a computer-readable storage medium is provided, on which computer program instructions are stored, and when the computer program instructions are executed by a processor, the environmental monitoring device calibration method is implemented.
[0090] The present invention may be a method, an apparatus, a system and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for executing various aspects of the present invention.
[0091] It should be understood by those skilled in the art that the embodiments of the present invention described above and shown in the accompanying drawings are only examples and do not limit the present invention. The purpose of the present invention has been fully and effectively achieved. The functional and structural principles of the present invention have been demonstrated and explained in the embodiments, and the embodiments of the present invention may be deformed or modified in any way without departing from the principles.
[0092] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for calibrating environmental monitoring equipment, It is characterized in that include: At the beginning of the current detection cycle, environmental samples are collected and saved; Obtain historical content information of multiple substances in environmental samples at the beginning of the previous detection cycle; Using an environmental monitoring spectral device, testing environmental samples at the beginning of a previous testing cycle to obtain comparative spectral data, and inputting the comparative spectral data into a content prediction model to obtain comparative content information of multiple substances; Determining whether the environmental monitoring spectroscopy device needs to be calibrated in a current detection cycle according to the comparative content information and the historical content information; If the environmental monitoring spectral device needs to be calibrated in the current detection cycle, the content prediction model is trained according to the comparative content information, the historical content information and the preset calibration sample to obtain the trained content prediction model and the calibrated environmental monitoring spectral device; Determine the duration of the current detection cycle according to the comparison content information, the historical content information, and the environmental sample at the start time of the current detection cycle; Performing environmental monitoring during the current detection cycle according to the calibrated environmental monitoring spectroscopy device, the trained content prediction model, and the duration of the current detection cycle; Moving the environmental sample at the start time of the previous detection cycle to the waiting recovery area and recovering it; Determining whether the environmental monitoring spectroscopy device needs to be calibrated in a current detection cycle according to the comparison content information and the historical content information includes: Determine, according to the comparative content information and the historical content information, a first type set of multiple substances corresponding to the historical content information and a second type set of multiple substances corresponding to the comparative content information; Determine an intersection of the first type set and the second type set; Determine a third type set included in the first type set but not included in the intersection; Determine a fourth type set included in the second type set but not included in the intersection; Determining whether the environmental monitoring spectroscopy device needs to be calibrated in a current detection cycle according to the comparison content information, the historical content information, the intersection, the third type set, and the fourth type set; If the environmental monitoring spectral device needs to be calibrated in the current detection cycle, the content prediction model is trained according to the comparative content information, the historical content information and the preset calibration sample to obtain the trained content prediction model and the calibrated environmental monitoring spectral device, including: Detecting sample spectrum data of the preset calibration sample by an environmental monitoring spectrum device; Inputting the sample spectral data into the content prediction model to obtain sample content information of multiple substances in the preset calibration sample; Determining a loss function of the content prediction model according to the sample content information, the annotation information of the preset calibration sample, the comparative content information and the historical content information; According to the loss function, the content prediction model is trained to obtain a trained content prediction model; According to the trained content prediction model, obtaining the calibrated environmental monitoring spectroscopy device; Determining a loss function of the content prediction model according to the sample content information, the annotation information of the preset calibration sample, the comparative content information and the historical content information includes: According to the formula Determine the loss function Loss of the content prediction model, where Ω 1 is a first type set, Ω 2 is the second type set, Ω 1 ∪Ω 2 is the union of the first type set and the second type set, count(Ω 1 ∪Ω 2 ) is the total number of types in the union of the first type set and the second type set, C 1,s is the historical content information of the s-th type of substance in the union of the first type set and the second type set, C 2,s is the comparative content information of the s-th type of substance in the union of the first type set and the second type set, ΔT is the duration of the previous detection cycle, T T is the duration of the preset monitoring period, C ET is the sensitivity error, Ω 5 is a fifth type set of multiple substances contained in the preset calibration sample determined according to the annotation information, Ω 6 is a sixth type set of multiple substances contained in the preset calibration sample determined according to the sample content information, Ω 5 ∪Ω 6 is the union of the fifth type set and the sixth type set, count(Ω 5 ∪Ω 6 ) is the total number of types in the union of the fifth type set and the sixth type set, C b,t is the content information of the substance of the tth type in the union of the fifth type set and the sixth type set determined according to the annotation information, C y,t is the sample content information of the t-th type of substance in the union of the fifth type set and the sixth type set, ΔT L is the time interval between the last training of the content prediction model and the start of the current detection cycle, s≤count(Ω 1 ∪Ω 2 ), t≤count(Ω 5 ∪Ω 6 ), and s and t are both positive integers.
2. The environmental monitoring equipment calibration method according to claim 1, It is characterized in that Determining whether the environmental monitoring spectroscopy device needs to be calibrated in a current detection cycle according to the comparison content information, the historical content information, the intersection, the third type set, and the fourth type set includes: According to the formula Obtain the weighted error E of the environmental monitoring spectrum equipment in two adjacent detection cycles, where C 1,i is the historical content information of the i-th type of substance in the intersection, C 2,i is the comparative content information of the i-th type of substance in the intersection, I is the total number of types in the intersection, C 3,j is the historical content information of the jth type of substance in the third type set, C ET is the sensitivity error, n 3 is the total number of types in the third type set, C 4,k is the content information of the kth type of substance in the fourth type set, n 4 is the total number of types in the fourth type set, if is the conditional function, i≤I, j≤n 3 , k≤n 4 , and i, j, k, I, n 3 and n 4 All are positive integers; When the weighted error is greater than or equal to a preset error threshold, it is determined that the environment monitoring spectrum device needs to be calibrated in the current detection cycle.
3. The environmental monitoring equipment calibration method according to claim 1, It is characterized in that Determining the duration of the current detection cycle according to the comparison content information, the historical content information, and the environmental sample at the start time of the current detection cycle includes: The measured content information of various substances in the environmental sample at the beginning of the current detection cycle is obtained through the calibrated environmental monitoring spectrum equipment; Determining a substance content variation coefficient according to the measured content information and the historical content information; Determining an environmental impact coefficient according to the comparative content information and the historical content information; The duration of the current detection cycle is determined according to the substance content variation coefficient and the environmental impact coefficient.
4. The environmental monitoring equipment calibration method according to claim 3, It is characterized in that Determining the duration of the current detection cycle according to the substance content variation coefficient and the environmental impact coefficient includes: According to the formula ΔT N =ΔT-if(H×M≥ε,HMΔT,-HMΔT) Determine the duration of the current detection cycle ΔT N ,in, is the environmental impact coefficient, is the coefficient of variation of the substance content, Ω 1 is a first type set, Ω 2 is the second type set, Ω 1 ∪Ω 2 is the union of the first type set and the second type set, count(Ω 1 ∪Ω 2 ) is the total number of types in the union of the first type set and the second type set, C 1,s is the historical content information of the s-th type of substance in the union of the first type set and the second type set, C 2,s is the comparative content information of the s-th type of substance in the union of the first type set and the second type set, ΔT is the duration of the previous detection cycle, Ω 7 is the seventh type set of various substances corresponding to the measured content information, Ω 1 ∪Ω 7 is the union of the first type set and the seventh type set, count(Ω 1 ∪Ω 7 ) is the total number of types in the union of the first type set and the seventh type set, C 1,h is the historical content information of the hth type of substance in the union of the first type set and the seventh type set, C c,h is the measured content information of the hth type of substance in the union of the first type set and the seventh type set, s≤count(Ω 1 ∪Ω 2 ), h≤count(Ω 1 ∪Ω 7 ), and s and h are both positive integers.
5. A system for performing an environmental monitoring device calibration as claimed in any one of claims 1 to 4, It is characterized in that include: The environmental sample module is used to collect and save environmental samples at the beginning of the current detection cycle; A historical content information module is used to obtain historical content information of multiple substances in environmental samples at the beginning of the previous detection cycle; A comparative content information module is used to detect the environmental sample at the start time of the previous detection cycle through an environmental monitoring spectrum device to obtain comparative spectrum data, and input the comparative spectrum data into a content prediction model to obtain comparative content information of multiple substances; A calibration module is used to determine whether the environmental monitoring spectroscopy device needs to be calibrated in the current detection cycle according to the comparison content information and the historical content information; A prediction model and spectral device module, for training the content prediction model according to the comparative content information, the historical content information and a preset calibration sample to obtain a trained content prediction model and a calibrated environmental monitoring spectral device if the environmental monitoring spectral device needs to be calibrated in the current detection cycle; A duration module, used to determine the duration of the current detection cycle according to the comparison content information, the historical content information, and the environmental sample at the start time of the current detection cycle; An environmental monitoring module, configured to perform environmental monitoring within the current detection cycle according to the calibrated environmental monitoring spectral device, the trained content prediction model and the duration of the current detection cycle; The recycling module is used to move the environmental sample at the start time of the previous detection cycle to the waiting recycling area and recycle it.
6. An environmental monitoring equipment calibration device, It is characterized in that include: processor; A memory for storing processor-executable instructions; wherein the processor is configured to call the instructions stored in the memory to execute the method according to any one of claims 1 to 4.
7. A computer-readable storage medium, It is characterized in that Computer program instructions are stored thereon, and when the computer program instructions are executed by a processor, the method according to any one of claims 1 to 4 is implemented.
Citation Information
Patent Citations
Online Raman spectrometer correction device and correction method thereof
CN102121902A
Environment monitoring method and system
CN111414586A
Environment monitoring method, system and platform and storage medium
CN113138865A
