Search

CN-121998527-A - Medicine cold chain logistics quality early warning method

CN121998527ACN 121998527 ACN121998527 ACN 121998527ACN-121998527-A

Abstract

The invention belongs to the technical field of medicine cold-chain logistics quality early warning, and particularly relates to a medicine cold-chain logistics quality early warning method, which comprises the steps of constructing a medicine quality full-link early warning index system and preprocessing specific index data by adopting normalization; the method comprises the steps of carrying out weight calculation on all early warning indexes by adopting a combined weighting method, introducing an SG filter and an SSA algorithm to optimize BiLSTM model, obtaining an SG-SSA-BiLSTM model, training the model, and carrying out warning condition prediction on the whole medicine quality cold chain logistics and all links of the medicine quality cold chain logistics through the model after optimization training. Compared with the prior art, the method has the advantages of highest prediction precision, more stable performance and more accurate early warning result, and can effectively reflect the real quality safety condition of the medicine, thereby helping management personnel predict and eliminate risks in advance, and having important practical significance for guaranteeing the quality safety of the medicine.

Inventors

  • WANG ZIXU
  • YANG WEI
  • Lv Danyang

Assignees

  • 陕西科技大学

Dates

Publication Date
20260508
Application Date
20260121

Claims (6)

  1. 1. The medicine cold-chain logistics quality early warning method is characterized by comprising the following steps of: 1) Constructing a medicine quality full-link early warning index system, and preprocessing specific index data by adopting normalization; 2) Performing weight calculation on all early warning indexes by adopting a combined weighting method; 3) Introducing an SG filter and an SSA algorithm to optimize BiLSTM models, obtaining an SG-SSA-BiLSTM model and training the model; 4) And predicting the medical quality cold chain logistics overall and the police condition of each link of the medical quality cold chain logistics through the model after the optimization training is completed.
  2. 2. The method for early warning of the quality of a medical cold chain logistics according to claim 1, wherein the medical quality full-link early warning index system in the step 1) specifically comprises 5 primary indexes and 22 secondary indexes; The first-level index comprises a production link, a verification link, a storage link, a transportation link and a use link; The secondary indexes of the production link comprise the quality of the medical raw materials and auxiliary materials, the level of medical production equipment, the level of personnel operation standardization and the level of medical production technology; the secondary indexes of the verification link comprise PH value, bacterial endotoxin, high molecular protein, phenol or metacresol content, zinc content, insoluble particle content and sterile detection; the secondary index of the storage link stores temperature, storage humidity and storage equipment level, and medical validity period checking frequency and expired medical destruction; The secondary indexes of the transportation link comprise transportation temperature, transportation equipment level, transportation humidity and transportation road conditions; The secondary indexes of the using link comprise the professional level of doctors and the standardized level of medication.
  3. 3. The method for pre-warning the quality of a pharmaceutical cold-chain logistics according to claim 1, wherein the combined weighting method in the step 2) is specifically a hierarchical analysis method-entropy method, and has the following formula Wherein, W Sj is the subjective weight of the jth index, W Oj is the objective weight of the jth index, and k is the coefficient, and the value range of k is (0, 1).
  4. 4. The method of claim 1, wherein the SG-SSA-BiLSTM model of step 3) comprises an input layer, a double-layer BiLSTM layer structure, a full-connection layer and an output layer, and the activation function is ReLU.
  5. 5. The method of claim 1, wherein the SG filter introduced in the step 3) is used for denoising the medicine quality evaluation value sequence, n data points are continuously selected from the first data point to the last data point as a window width, a least square method is used as a fitting polynomial, and windows are sequentially and circularly selected to perform data fitting until the last window completes data fitting.
  6. 6. The method for pre-warning the quality of a pharmaceutical cold-chain logistics according to claim 1, wherein the SSA algorithm optimization parameters in the step 3) comprise BiLSTM layers of neuron numbers, full-connection layer neuron numbers, learning rate, batch processing amount and iteration times; The update formula is: Wherein: -the j-th dimension position of the ith sparrow in the T-th iteration, T-the maximum iteration number of the population, alpha-a uniform random number, Q-a standard normal distribution random value, L-a matrix with the size of 1x d which is all 1, R2-a sparrow population early warning value and ST-a sparrow population safety threshold.

Description

Medicine cold chain logistics quality early warning method Technical Field The invention belongs to the technical field of medical cold-chain logistics quality early warning, and particularly relates to a medical cold-chain logistics quality early warning method. Background The global medicine cold chain logistics demand is greatly increased, the development prospect is wide, but the industry frequently generates quality safety accidents, the root causes are that raw materials are poor, storage and transportation equipment is insufficient, personnel are poor in professional, and the problems of medicine deterioration, fake and inferior medicine inflow and the like are easily caused by uncontrolled environments such as superimposed humiture and the like, and the safety tracing and dynamic monitoring become key. The short board in the prior art is prominent in that most enterprises rely on manual recording of temperature and humidity and other data, the defects of discontinuity and inaccuracy exist, and in that some large enterprises apply RFID, GPS and other technologies, but due to complex links and huge sensor data, risks of scattered storage and easy tampering and leakage are faced, an industrial anti-counterfeiting traceability mechanism is imperfect, responsibility identification and recall are difficult, most of existing traceability code data are enterprise concentrated monopoly, encryption deletion causes low reliability of consumer query information, and authenticity is difficult to distinguish. The technology innovation and equipment upgrading are needed to construct a safety early warning and credible traceability system so as to promote the high-quality development of a medicine cold chain. The traditional early warning research on the quality safety of the traditional Chinese medicine in the cold chain logistics is mainly focused on controlling and early warning the temperature and humidity in storage or transportation, evaluating the risk of the medicine cold chain logistics and the like, and the research on influencing the quality safety of the medicine in each link of the medicine cold chain logistics is rarely comprehensively considered. Therefore, research on the quality safety problem of the medicine cold chain logistics by using the advanced technology at present is imperative. Disclosure of Invention The invention aims to provide a medical cold-chain logistics quality early warning method which is used for solving the problems existing in the prior art and realizing comprehensive consideration early warning of the whole process of medical cold-chain logistics. In order to achieve the above purpose, the invention adopts the following technical scheme: A medicine cold chain logistics quality early warning method comprises the following steps: 1) Constructing a medicine quality full-link early warning index system, and preprocessing specific index data by adopting normalization; 2) Performing weight calculation on all early warning indexes by adopting a combined weighting method; 3) Introducing an SG filter and an SSA algorithm to optimize BiLSTM models, obtaining an SG-SSA-BiLSTM model and training the model; 4) And predicting the medical quality cold chain logistics overall and the police condition of each link of the medical quality cold chain logistics through the model after the optimization training is completed. Further, the medicine quality full-link early warning index system in the step 1) specifically comprises 5 primary indexes and 22 secondary indexes; The first-level index comprises a production link, a verification link, a storage link, a transportation link and a use link; The secondary indexes of the production link comprise the quality of the medical raw materials and auxiliary materials, the level of medical production equipment, the level of personnel operation standardization and the level of medical production technology; the secondary indexes of the verification link comprise PH value, bacterial endotoxin, high molecular protein, phenol or metacresol content, zinc content, insoluble particle content and sterile detection; the secondary index of the storage link stores temperature, storage humidity and storage equipment level, and medical validity period checking frequency and expired medical destruction; The secondary indexes of the transportation link comprise transportation temperature, transportation equipment level, transportation humidity and transportation road conditions; The secondary indexes of the using link comprise the professional level of doctors and the standardized level of medication. Further, the combined weighting method in the step 2) is specifically an analytic hierarchy process-entropy value method, and the following formula is adopted Wherein, W Sj is the subjective weight of the jth index, W Oj is the objective weight of the jth index, and k is the coefficient, and the value range of k is (0, 1). Further, in the SG-SSA-BiLSTM model in the step