CN-122020170-A - Oil painting preservation environment risk assessment method

CN122020170ACN 122020170 ACN122020170 ACN 122020170ACN-122020170-A

Abstract

A method for evaluating the environmental risk of oil painting preservation includes such steps as collecting environmental data, constructing a time sequence, preprocessing the environmental data to obtain environmental time sequence data, slicing the environmental time sequence data to obtain the characteristic samples for inputting the time sequence prediction model, constructing an improved time sequence prediction model, predicting the environmental parameters of oil painting based on the improved time sequence prediction model, and outputting the prediction result. According to the invention, through trend prediction and comprehensive evaluation of the environmental parameters, the passive response of environmental management is converted into early warning, and corresponding measures can be taken before the environmental abnormality has substantial influence on oil painting.

Inventors

Jue Zhihong
CAO HUAJIE
AN YONGKANG
LUO KAI
HAN FEI
KE YONGBIN
HAN YING
ZOU ZIJIAN
WANG CHONGHUA
ZHU YUHANG
CAI ZHIYONG
YANG RUYI

Assignees

淮阴工学院

Dates

Publication Date: 20260512
Application Date: 20260129

Claims (8)

1. A risk assessment method for oil painting preservation environment is characterized by comprising the following steps: collecting environment data and constructing a time sequence; Preprocessing the environmental data acquired in the first step to obtain environmental time sequence data, slicing the environmental time sequence data by adopting a sliding mode of a fixed time window, and constructing a characteristic sample for inputting a time sequence prediction model; Thirdly, constructing an improved time sequence prediction model, constructing the improved time sequence prediction model based on a long-short-term memory neural network LSTM, and improving a memory evolution mechanism of the oil painting preservation environment according to risk assessment requirements of the oil painting preservation environment; And fourthly, based on the prediction result output in the third step, modeling analysis is carried out on the long-term influence possibly generated by the oil painting material based on the predicted environmental condition, so that the early recognition and quantitative evaluation of the oil painting preservation environmental risk are realized, and the continuous and cumulative influence on the oil painting caused by the characterization environmental factors in the time dimension is focused during evaluation, so that the degradation process of aging, fading, cracking and pollution adsorption of the oil painting material is reflected more truly.
2. The oil painting preservation environment risk assessment method is characterized by comprising the following steps of firstly, collecting environment data, wherein a plurality of environment sensors are deployed in an oil painting display space and used for monitoring micro-environment parameters around the oil painting for a long time, the environment parameters collected by the environment sensors are collected according to a unified time standard, the environment sensors comprise environment sensors used for collecting temperature, relative humidity, illumination intensity, ultraviolet radiation intensity and air pollutant concentration, the temperature sensors are used for obtaining air temperature change conditions around the oil painting, the relative humidity sensors are used for reflecting water content change in the air, the illumination intensity sensors are used for monitoring irradiation intensity of visible light on the surface of the oil painting, the ultraviolet radiation sensors are used for monitoring ultraviolet radiation levels which possibly cause pigment fading and material aging, the air pollutant sensors are used for detecting pollutant concentrations with potential corrosion effects on oil painting materials in the air, and day and night identification data of the sensors are automatically generated according to collecting time information and are used for distinguishing environment states under natural illumination conditions and non-illumination conditions.
3. The method for evaluating the risk of oil painting preservation environment according to claim 1, wherein the specific operation mode of the first step is that the time sequence is constructed in the first step, and the specific operation mode is that the sampling time interval is set to be a fixed time interval Each sampling time corresponds to a group of environment state vectors It can be expressed as: (1); Wherein, the Indicating time of day Is used for the temperature control of the air conditioner, Indicating the relative humidity of the sample, The intensity of the light is indicated and, Indicating the intensity of the ultraviolet radiation, Indicating the concentration of the contaminant in the air, Representing diurnal identification variables; Arranging environment state vectors obtained by continuous acquisition according to a time sequence to construct a multidimensional time sequence data set for forming an oil painting preservation environment : (2); Wherein, the Representing the total number of time steps for data acquisition.
4. The oil painting preservation environment risk assessment method according to claim 1, wherein the preprocessing of the environment data in the second step is performed in the following specific operation modes: Step 2.1, aiming at missing data, adopting a corresponding data filling strategy according to the missing condition, wherein the data filling strategy is to fill the missing data by adopting an interpolation mode based on time continuity when the missing time is short and the data of the front time point and the rear time point are complete; When an abnormal value is detected, adopting a mode of smooth substitution of adjacent time point data, sliding average correction or limitation in a reasonable interval to carry out correction processing; And 2.3, carrying out normalization processing on each environmental parameter, mapping the data with different dimensions to a unified numerical range, and improving the convergence speed and the prediction accuracy of model training. The environmental parameters are consistent in numerical scale.
5. The oil painting preservation environment risk assessment method according to claim 1, wherein the construction feature sample in the second step is specifically implemented by setting a time window length as I.e. each characteristic sample comprises a succession of Sequentially intercepting a plurality of groups of continuous environmental state data by sliding a time window on a time axis with a preset step length to form a characteristic sample set, wherein each characteristic sample is a set of the characteristic samples Can be expressed as: (3); Wherein, the Indicating time of day Is a multi-dimensional environmental parameter vector; the single characteristic sample contains environmental state information at the current moment, and integrates environmental change processes at a plurality of historical moments, so that the subsequent time sequence prediction model can fully learn the time dependency relationship and the change trend of environmental parameters.
6. The method for evaluating the risk of oil painting preservation environment according to claim 1, wherein the step three is characterized in that the environmental parameters of the oil painting are predicted, and in the time sequence prediction process, the model takes the environmental parameters in a plurality of continuous time steps as input to describe the change process of the oil painting preservation environment in a period of time; Is set at a time step The environmental parameter vector of the moment is The model can sense the history state and the change direction of the environmental parameters at the same time, thereby providing a sufficient information basis for the subsequent prediction And hidden state Storing and updating the history information of the environmental parameters to realize modeling of the long-term evolution rule of the environment, and at each time step The LSTM network selectively memorizes the history information through an input gate, a forget gate and an output gate, and the basic updating process is expressed as follows: (4); Wherein, the Representing time steps Is hidden in the first layer; Representing time steps History cell state of (a); The outputs of the forget gate and the input gate are respectively; Representing candidate memory information generated jointly by the current environmental parameter input and the historical hidden state; Indicating time of day The memory unit state of the oil painting storage environment is used for storing long-term evolution information of the oil painting storage environment in a time dimension; 、 Is a trainable parameter matrix and a bias term; And (3) with Representing the Sigmoid function and the hyperbolic tangent function respectively, Representing an element-by-element multiplication operation.
7. The method for evaluating the risk of oil painting preservation environment according to claim 6, wherein the method for constructing the improved time series prediction model in the third step comprises the steps of introducing an environmental change cumulative amount in a time dimension based on an LSTM network model for representing the duration of environmental change in a period of time, and then the environmental change cumulative amount The definition is as follows: (5); Wherein, the Is that The vector of the environmental parameters at the moment in time, Is that The cumulative amount of the environmental change expressed by the formula (5) is used for distinguishing the situation that the environmental change is sporadically and severely changed in a short time, and the environmental drift is continuous and slow in a long time but consistent in direction; To characterize the stability of the direction of environmental change, an environmental change consistency factor is introduced : (6); Wherein the consistency factor The device is used for measuring whether the environmental change has similar change directions in adjacent time periods, wherein the factor takes a larger value when the environmental change shows long-term unidirectional drift, and takes a smaller value when the environmental change direction is frequently reversed; comprehensively considering the cumulative amplitude, the change persistence and the directional stability of environmental changes, constructing a memory attenuation regulating factor, and expressing the memory attenuation regulating factor as follows: (7); Wherein, the The mapping function is used for limiting the regulating factor within a preset range; Is a weight parameter; Is a regulating factor and is used for dynamically controlling the retention degree of the history memory; In order to make the memory updating process more in line with the physical characteristics of the long-term evolution of the oil painting preservation environment, an environment evolution driven memory attenuation regulating factor is introduced on the basis of the basic structure Further adjusting the retention strength of the history memory information, and updating the memory unit Hidden state Expressed as: (8); By the formula (8), the model can strengthen the memory of the historical environment state when the environment changes slowly and the trend is stable, automatically reduce the influence of the historical information on the current prediction when the environment changes suddenly or fluctuates disorderly, and improve the stability and physical rationality of the prediction result; based on updated memory cell And hidden state The model outputs the environmental parameter prediction result of the future time step Expressed as: (9); Wherein, the Is a bias term; The prediction result not only reflects the instantaneous change trend of the environmental parameter, but also implies the long-term evolution characteristic of the environment, and provides high-reliability prediction input for the accumulated damage evaluation and the environmental risk detection of the subsequent oil painting.
8. The method for evaluating the risk of oil painting preservation environment according to claim 1, wherein the specific operation mode of the fourth step is as follows: step 4.1, according to the output result of the step three, obtaining a predicted environment parameter sequence arranged in time sequence in a future predicted time interval: (10); wherein each predicted environmental parameter vector is defined as: (11); The predicted environment parameter sequence is used for describing the overall environment evolution trend of the oil painting storage space in a future period of time, rather than an isolated single-moment state, and provides a time-continuous input basis for subsequent damage modeling; step 4.2 quantitatively describing the deviation of the predicted environmental parameter from the ideal preservation condition for any environmental parameter Constructing a deviation function thereof Expressed as: (12); Wherein, the A target value indicating a long-term storage recommendation of the oil painting; representing the normal fluctuation range allowed by the environmental parameter; Is the first A deviation function value of each predicted environmental parameter; Is the predicted first The deviation function reflects the relative deviation degree of the environmental parameter relative to the ideal state and provides a unified measurement basis for different dimension parameters; step 4.3 introducing a nonlinear environmental stress intensity function for any environmental parameter The environmental stress intensity is defined as: (13); Wherein, the Is the first Environmental stress intensity values of the individual predicted environmental parameters; Is the first A deviation function value of each predicted environmental parameter; the index is used for describing the sensitivity difference of the oil painting to different environmental factors, so that the model can show nonlinear turning characteristics between slow accumulation and fast deterioration; step 4.4, defining transient damage increment based on the multidimensional environmental stress intensity as follows: (14); Wherein, the Is the first Instantaneous injury increment value of individual, weight coefficient The instantaneous damage increment reflects the influence of a single time step only and cannot directly represent the integral preservation state of the oil painting; step 4.5, constructing a cumulative damage evolution model based on the time memory effect, and defining the cumulative damage evolution model in time The accumulated damage index of the oil painting at the moment is The evolution relationship is as follows: (15); Wherein, the To a cumulative damage index, to characterize the extent of potential degradation of the oil painting under prolonged environmental exposure; for regulating the duration of historical injury in current stage, the accumulated injury at the last moment The current state is continuously influenced in a certain proportion without automatically disappearing in the time advancing process; step 4.6, dynamically associating the damage memory retention factor with the predicted environment change amplitude: (16); Wherein, the Is a non-negative adjustment coefficient for controlling the influence intensity of the predicted environmental change amplitude on the damage memory retention factor when When the environmental change is larger, the inhibition effect of the environmental change on the history injury memory is more obvious, when Less, the sensitivity of impaired memory to environmental changes is reduced; and 4.7, carrying out grading judgment on the risk of the oil painting preservation environment based on the change condition of the accumulated damage index in the prediction time interval, and specifically, defining the risk grade as follows: (17); Wherein, the Respectively representing a lower limit and an upper limit of the cumulative damage index; The system is used for representing the risk level and comprehensively reflecting the potential threat degree of the environmental condition to the long-term preservation safety of the oil painting in a future period, and when the predicted environmental risk level reaches the medium risk or the high risk, the system generates a targeted environmental regulation suggestion or control instruction according to the type of the environmental parameter of a risk source, wherein the targeted environmental regulation suggestion or control instruction comprises measures such as temperature and humidity regulation, illumination limitation, ultraviolet shielding, air purification and the like.

Description

Oil painting preservation environment risk assessment method Technical Field The invention relates to the technical field of cultural relic environment management and data processing, in particular to an oil painting preservation environment risk assessment method. Background The oil painting is composed of canvas, primer, pigment layer, protective coating and other materials, and each material can gradually undergo irreversible degradation phenomena such as aging, fading, cracking or pollution adsorption under the long-term action of environmental factors such as temperature, relative humidity, illumination, ultraviolet radiation, air pollutants and the like. At present, the existing art gallery oil painting preservation environment monitoring mostly adopts a real-time monitoring mode based on a fixed threshold value, environmental parameters are continuously collected by arranging environmental sensors such as temperature, humidity, illumination, ultraviolet radiation and the like in a display space or a storehouse, and the real-time monitoring value is compared with a preset safety threshold value. When a certain environmental parameter exceeds a threshold range, the system triggers an alarm to prompt a manager to take corresponding regulation and control measures. However, this manner of threshold-based environmental monitoring essentially focuses only on the instantaneous state of the environmental parameter at a single moment, and its risk judgment logic is based on a binary judgment of "whether the threshold is exceeded" so as to hardly reflect the continuous change process of the environmental condition in the time dimension. When environmental parameters are in a state close to a threshold value for a long time but not actually exceeding the standard, the method does not trigger any risk prompt, but oil painting materials can be continuously and negatively influenced in the 'sub-safe' environment, and finally irreversible accumulated damage is caused. In order to make up for the defects of the fixed threshold method, a statistical analysis or an empirical rule method is introduced into the environment monitoring process of a part of art museums, such as carrying out sliding average, daily average or periodic statistical analysis on environment parameters, and comprehensively judging the environment state by combining with manual experience, so that the influence caused by short-time fluctuation can be smoothed to a certain extent, but the analysis process is mostly based on linear statistical characteristics, and the nonlinear response characteristic of oil painting materials to environment changes is difficult to be described. Meanwhile, the method relies on manual experience for risk interpretation, is high in subjectivity, is difficult to normalize and quantify an evaluation result, and still lacks the capability of predicting the future environment change trend. And part of the environment management system adopts a rule base or expert system mode, and performs risk identification on different environment parameter combination states by presetting a large number of 'if-then' judgment rules. The system has certain applicability in a specific scene, but the rule system is highly dependent on priori knowledge, and is difficult to cover a complex change mode in the long-term evolution process of environmental conditions. When multi-factor coupling or slow drift exists between environment parameters, the rule base is often difficult to identify potential risks in time, and the system expansion and maintenance cost is also obviously increased along with the increase of the number of the rules. Disclosure of Invention In order to solve the technical problem that the long-term potential risk is difficult to identify in time due to the fact that only the instantaneous state of environmental parameters at a single moment is concerned in the oil painting preservation environment risk assessment process, the evolution trend of environmental conditions along with time and the accumulative property and the irreversibility of oil painting material degradation process are ignored, the technical scheme provides the oil painting preservation environment risk assessment method, so that the oil painting preservation environment risk assessment is converted into an active early warning mode from a traditional passive warning mode, the risk identification accuracy, assessment stability and the scientificity and reliability of protection decision in the oil painting long-term preservation process are obviously improved, and the technical problem can be effectively solved. The invention is realized by the following technical scheme: a risk assessment method for oil painting preservation environment comprises the following steps: collecting environment data and constructing a time sequence; Preprocessing the environmental data acquired in the first step to obtain environmental time sequence data, slicing the environmen