Search

CN-122016005-A - Multi-mode liquid level detection method and system for pipetting operation

CN122016005ACN 122016005 ACN122016005 ACN 122016005ACN-122016005-A

Abstract

The invention relates to a multi-mode liquid level detection method and system for pipetting operation, and belongs to the technical field of laboratory automation liquid treatment. The method comprises the steps of obtaining pipetting operation detection parameters, screening detection distribution points in a pipetting stroke domain through a time sequence change point detection algorithm to obtain pipetting operation feature sequences, constructing a pipetting response sensing matrix, calculating equipment state monitoring weights corresponding to detection modes according to response correlation values to generate an equipment state monitoring scheme, modeling a liquid level detection process, calculating quantiles of pipetting detection equipment in different operation states through a quantile statistics evaluation algorithm, outputting liquid level detection results, adjusting abnormal operation offset parameters based on a response feedback consistency mechanism by the liquid level detection results, mapping liquid level detection signals to a pipetting operation-response state chain, correcting contribution coefficients of all detection modes, and updating the equipment state monitoring scheme.

Inventors

  • FENG RENBING

Assignees

  • 上海硕控电子科技有限公司

Dates

Publication Date
20260512
Application Date
20260203

Claims (12)

  1. 1. A multi-modal fluid level detection method for pipetting operations comprising: S1, acquiring pipetting operation detection parameters, identifying candidate liquid level response points through a time sequence change point detection algorithm, generating a liquid level detection signal, constructing a liquid level pipetting signal window based on the liquid level detection signal, and introducing a multi-mode detection attenuation factor to calculate a response correlation value corresponding to the current pipetting operation to obtain a pipetting operation feature sequence; S2, constructing a pipetting response sensing matrix, acquiring time sequence evolution relations of pipetting operation feature sequences in different pipetting stages, inquiring equipment state monitoring weights corresponding to the current pipetting stage according to the response association values, and generating an equipment state monitoring scheme; S3, executing the equipment state monitoring scheme, modeling a liquid level detection process, calculating the quantiles of the pipetting detection equipment in different operation states through a quantile statistics evaluation algorithm, carrying out state sensitive correction on normal liquid level response in a liquid level detection signal and abnormal response caused by equipment state fluctuation based on the quantiles, and outputting a liquid level detection result; And S4, adjusting abnormal operation offset parameters in the liquid level detection result based on a response feedback consistency mechanism, mapping a liquid level detection signal to a pipetting operation-response state chain, correcting a contribution coefficient of the equipment state monitoring weight, and updating an equipment state monitoring scheme.
  2. 2. The method of claim 1, wherein the method for constructing the liquid level pipetting signal window comprises the steps of identifying candidate liquid level response points in a pipetting stroke domain through a time sequence change point detection algorithm in pipetting operation detection parameters, extracting multi-mode signal features by taking the candidate liquid level response points as the center, introducing multi-mode detection attenuation factors, covering a pre-response stage before liquid level contact, a liquid level crossing transient state and a recovery stage after liquid level stabilization, and obtaining the liquid level pipetting signal window.
  3. 3. The method according to claim 1, wherein the multi-mode detection attenuation factor is a piecewise linear attenuation function using a time difference between a detection mode response occurrence time and a current pipetting operation evaluation time and an operation travel offset as independent variables, and for each detection mode, substituting the time difference and the operation travel offset as independent variables into the corresponding piecewise linear attenuation function, and weighting the detection mode response characteristic to obtain a response correlation value under the current pipetting operation.
  4. 4. The method of claim 1, wherein the method for generating the pipetting operation feature sequence comprises the steps of dynamically weighting and fusing the response points based on the multi-mode detection attenuation factors, organizing the fused response points according to pipetting operation time sequence to form a time sequence response lag signal, mapping the liquid level response in each stage to a discrete response vector caused by operation disturbance extracted by a node of a corresponding operation stage according to a preset operation stage division rule, and obtaining the pipetting operation feature sequence.
  5. 5. The method of claim 2, wherein the method for constructing the pipetting response sensing matrix is characterized by dividing the characteristic sequence according to pipetting strokes and time dimensions based on the pipetting characteristic sequence, and constructing the pipetting response sensing matrix of the phase association and time sequence evolution relation of the liquid level response characteristics in the pipetting process by taking different pipetting phases and corresponding multi-mode response characteristics as matrix dimensions.
  6. 6. The method of claim 5, wherein the calculating method of the equipment state monitoring weight is characterized in that the pipetting response sensing matrix is layered according to pipetting operation stages, response consistency indexes and stability indexes under different pipetting operation stages are detected, the response consistency indexes are statistics of liquid level responses of different detection modes in time sequence positions, response amplitudes and change trends in the same pipetting operation stage, the stability indexes are measurement values of liquid level response fluctuation amplitudes of detection modes in continuous pipetting operation or adjacent pipetting operation stages, and the response association values are used as adjustment contribution references, and the consistency indexes and the stability indexes are combined, weighted and fused to obtain the equipment state monitoring weight.
  7. 7. The method of modeling a liquid level detection process according to claim 4, wherein the liquid level detection process is characterized in that the liquid level detection process is modeled by taking the liquid level operation characteristic sequence as an operation behavior input, taking the liquid level response sensing matrix as a liquid level response structure, introducing the equipment state monitoring weight, restraining response contribution caused by equipment state fluctuation, modeling a multi-mode liquid level response in a subsection mode according to a liquid level operation stage, intercepting a liquid level detection signal in a stage by taking a time interval of a corresponding stage in the liquid level operation characteristic sequence as a constraint in each liquid level operation stage, obtaining an instantaneous amplitude characteristic and a time sequence distribution density characteristic of the liquid level detection signal in the stage, parametrically modeling the time sequence distribution density characteristic, extracting the discrete degree and tail distribution characteristic of the liquid level detection signal in the liquid level operation stage, and constructing a liquid level detection response model.
  8. 8. The method according to claim 2, wherein the quantitive statistical evaluation algorithm performs joint statistics on probability distribution forms of the liquid level detection signals in the corresponding pipetting operation phases based on the liquid level detection response model in the liquid level detection process, calculates quantiles of the liquid level detection signals in the pipetting operation phases according to a preset quantile proportion, and quantifies the offset degree of a normal liquid level response interval of the liquid level detection signals in the coincidence phase response model and an abnormal fluctuation interval deviating from the response model.
  9. 9. The method of claim 4, wherein the liquid level detection result is output by taking the equipment state monitoring weight as a state sensitive adjustment factor, performing joint evaluation on the instantaneous amplitude characteristic and the time sequence distribution characteristic of the liquid level detection signal in the current pipetting operation stage, finding out a bit-dividing statistical reference interval in a corresponding operation state, eliminating detection response points falling into abnormal bit-dividing intervals according to the bit-dividing statistical reference interval, and outputting the liquid level detection result of the liquid level response in the candidate interval in the time dimension and the travel dimension.
  10. 10. The method of claim 1, wherein the responsive feedback consistency mechanism comprises a liquid level response map and a feedback correction adjustment; The liquid level response mapping maps the liquid level detection signals subjected to state sensitivity correction with the operation behavior characteristics of the corresponding pipetting operation stages, calculates the deviation index of the liquid level response of each operation stage from the expected liquid level response mode, and obtains a pipetting operation-liquid level response mapping chain; And the feedback correction adjustment reversely maps the deviation index of the liquid level response to an abnormal operation deviation parameter space according to the weighted correction result of the response point, dynamically corrects the liquid level response which is excessively deviated or does not accord with the continuity constraint, recalculates the contribution of the liquid level response by combining the corrected liquid level detection result, and updates the contribution to the equipment state monitoring scheme and the liquid level detection signal mapping chain.
  11. 11. The method according to claim 1, wherein the pipetting operation-response state chain correction contribution coefficient is characterized in that the method comprises the steps of mapping a liquid level response signal in each pipetting operation stage to a corresponding operation stage node, calculating a liquid level response deviation index under the node, carrying out joint initialization on an original contribution coefficient and a corresponding equipment state monitoring weight, reducing the contribution coefficient of a low stability mode through a weighted attenuation function for a chain node with larger deviation, and applying the corrected contribution coefficient to the liquid level response mapping chain to generate a new pipetting operation-response state chain.
  12. 12. A multi-modal fluid level detection system for pipetting operations for performing the method of any one of claims 1-11, comprising: the pipetting operation feature sequence construction module is used for acquiring pipetting operation detection parameters, identifying candidate liquid level response points through a time sequence change point detection algorithm, generating a liquid level detection signal, constructing a liquid level pipetting signal window based on the liquid level detection signal, introducing a multi-mode detection attenuation factor to calculate a response association value corresponding to the current pipetting operation, and obtaining a pipetting operation feature sequence; the equipment state monitoring scheme generation module is used for constructing a pipetting response sensing matrix, acquiring time-sequence evolution relations of pipetting operation feature sequences in different pipetting stages, inquiring equipment state monitoring weights corresponding to the current pipetting stage according to the response correlation values and generating an equipment state monitoring scheme; the detection modeling sensitivity correction module is used for executing the equipment state monitoring scheme, modeling a liquid level detection process, calculating the quantile of the pipetting detection equipment in different operation states through a quantile statistics evaluation algorithm, carrying out state sensitivity correction on the normal liquid level response in the liquid level detection signal and the abnormal response caused by equipment state fluctuation based on the quantile, and outputting a liquid level detection result; And the response feedback consistency correction module is used for adjusting abnormal operation offset parameters in the liquid level detection result based on a response feedback consistency mechanism, mapping a liquid level detection signal to a pipetting operation-response state chain, correcting the contribution coefficient of the equipment state monitoring weight and updating the equipment state monitoring scheme.

Description

Multi-mode liquid level detection method and system for pipetting operation Technical Field The invention belongs to the technical field of laboratory automation liquid treatment, and particularly relates to a multi-mode liquid level detection method and system for pipetting operation. Background Pipetting is a key element in laboratory liquid handling, and its accuracy of operation directly affects the reliability and reproducibility of experimental results. In the prior art, liquid level detection mainly relies on a single-mode sensor, such as an optical sensor or a capacitive sensor, to monitor the position of the liquid surface in real time. However, the single-mode detection has various limitations that the optical method is obviously influenced by liquid color, transparency and ambient light, and is difficult to adapt to complex experimental conditions, and the capacitance or pressure sensor is sensitive to liquid vibration and bubble interference and easily generates noise signals, so that the liquid level judgment deviation is caused. In addition, the traditional liquid level detection method generally adopts a fixed threshold value or a simple filtering algorithm, so that effective analysis on time sequence change and phase characteristics of a liquid level response signal is lacking, and abnormal response caused by actual liquid level change and equipment state fluctuation is difficult to distinguish. During pipetting operations, small changes in liquid volume and discontinuities in pipette operation can cause the liquid level signal to exhibit non-stationary, time-varying characteristics, increasing the complexity of liquid level detection. The existing method is difficult to realize fine modeling of the operation stages, and lacks a hierarchical response analysis means for different operation stages. Particularly in the aspect of multi-mode signal fusion, the traditional technology lacks an effective weighting strategy, and cannot fully utilize complementary information of multi-source signals such as optics, capacitance, pressure and the like, so that the accuracy and stability of liquid level detection are affected. Meanwhile, equipment state fluctuation such as vibration of a pipette, mechanical abrasion or environmental disturbance can introduce transient interference, and detection reliability is further reduced. In the prior art, the dynamic coupling relation between the liquid level response and the equipment state is not considered in most cases, and a self-adaptive correction mechanism based on response feedback is also lacking, so that the liquid level detection result is easy to be abnormal, and the experimental requirements of high precision and high reliability are difficult to meet. Furthermore, existing systems typically provide only a single value or signal at the level detection result output, and lack a periodic response map and continuity constraints to provide usable closed loop information for automated control or equipment status monitoring. Therefore, how to realize high-precision, multi-mode and staged liquid level detection in the pipetting operation process, realize abnormal response correction by combining equipment state monitoring and generate continuous and reliable liquid level detection results is a key problem to be solved by the current laboratory liquid treatment technology. Disclosure of Invention In order to solve the above-mentioned problems in the prior art, the present invention provides a multi-modal liquid level detection method for pipetting operations, The aim of the invention can be achieved by the following technical scheme: S1, acquiring pipetting operation detection parameters, identifying candidate liquid level response points through a time sequence change point detection algorithm, generating a liquid level detection signal, constructing a liquid level pipetting signal window based on the liquid level detection signal, and introducing a multi-mode detection attenuation factor to calculate a response correlation value corresponding to the current pipetting operation to obtain a pipetting operation feature sequence; S2, constructing a pipetting response sensing matrix, acquiring time sequence evolution relations of pipetting operation feature sequences in different pipetting stages, inquiring equipment state monitoring weights corresponding to the current pipetting stage according to the response association values, and generating an equipment state monitoring scheme; S3, executing the equipment state monitoring scheme, modeling a liquid level detection process, calculating the quantiles of the pipetting detection equipment in different operation states through a quantile statistics evaluation algorithm, carrying out state sensitive correction on normal liquid level response in a liquid level detection signal and abnormal response caused by equipment state fluctuation based on the quantiles, and outputting a liquid level detection result; And S4,