Prediction Method Of Radioactive Contamination In Nuclear Explosion Based On Data Assimilation

With the continuous advancement of nuclear technology,the number of countries possessing nuclear weapons in the world is increasing,which has resulted in a clear trend of nuclear confrontation and proliferation at the regional level,and the activation of nuclear terrorism has brought new threats to national security.In the event of a nuclear war or an act of nuclear terrorism,the misuse of nuclear weapons would bring about the destruction of all mankind.Among all the injuries caused by nuclear weapon explosions,radioactive contamination caused by the spread of radioactive particles into the environment with atmospheric movement is one of the longest lasting and most harmful injuries.In order to solve the practical problems of obtaining accurate source parameters and reducing the error of system model analysis in the actual process,it is necessary to address the trend of radioactive puff diffusion and radioactive contamination after nuclear weapon detonation on the ground or at low altitude.In this paper,a fast and stable data assimilation method with simple parameters for the prediction model of the radioactive contamination hazard trend of nuclear explosion is studied,which can provide high accuracy radioactive contamination prediction results with the measured data,and can provide fast and effective support for the command decision in a short time.The main work and conclusions of this paper are as follows:1.A model of nuclear explosion radioactive dispersion based on the sedimentation rate function was developed.In this paper,a calculation procedure for the prediction model of nuclear explosion radioactive puff dispersion with variable wind direction was established based on the sedimentation rate function,and the sedimentation rate model was improved by using the time division processing and coordinate conversion methods.The dose distribution of radioactive contamination on the ground with time can be simulated and calculated in the case of wind direction variation.To calculate the dispersion of radioactive contaminants from nuclear explosions and the prediction results of wind direction change.The prediction visualization of radioactive contamination hazard trend of nuclear explosion is realized by using R language.It also passed the validation of historical data,model comparison and general validation.The model can be used to predict the ground radiation area and radioactive contamination dose level assessment from nuclear explosion,and also provides a positive calculation model for radioactive contamination hazard prediction for the later data assimilation calculation.2.In order to solve the nuclear weapon explosion source term parameters,meteorological conditions are not clear,the identification process are complex and other problems.Combined with real-time monitoring data,by defining the fitness function,proposed a nuclear weapon source term parameters based on genetic algorithm and particle swarm optimization algorithm and other data assimilation methods identification method.After the optimization of model parameters and the verification of real historical data,the results show that both the particle swarm optimization algorithm and the genetic algorithm are able to identify the nuclear explosion source term parameters more ideally and maintain high prediction accuracy and precision after optimization.When the maximum number of iterations and population size of particle swarm optimization algorithm and genetic algorithm are the same,the running time and optimization accuracy of the particle swarm optimization algorithm are better than those of genetic algorithm.The study of this data assimilation method enriches the theory and method of radioactive particle dispersion prediction after nuclear weapon explosion,which is of great significance to the study of environmental radioactive particles and also lays the research foundation for the sequential data assimilation algorithm in the later paper.3.Based on the accurate source term parameters obtained by the data assimilation algorithm,the data assimilation algorithm combined with the nuclear weapon explosion radioactive contamination prediction model is used to create a Kalman filter algorithmbased radioactive contamination dose field data assimilation process by combining realtime monitoring data.The radioactive contamination prediction model is used to calculate the radiation dose level on the ground at a certain moment after the explosion and obtain the radioactive contamination dose rate distribution field at that moment.When monitoring data are available at that moment,the data are introduced into the assimilation calculation model.The main experimental results show that the data assimilation method can be used to optimize the prediction results of radioactive contamination of nuclear weapon detonation by combining multiple real-time monitoring data.The data assimilation algorithm improves the calculation accuracy and forecast accuracy of the nuclear explosion prediction model.Finally,the error estimation method of the data assimilation algorithm is studied,and the mechanism of the observation error is analyzed to analyze the uncertainty of the model and prediction values.4.In response to the needs of emergency management decision-making after a nuclear weapon attack and nuclear strike,as well as the practical problems of the nuclear weapon explosion radioactive contamination prediction model algorithm studied in this paper,such as low versatility in operation and visualization,and lack of software for rapid generation of prediction results in the real process,a data assimilation radiation hazard trend prediction system based on real-time monitoring data was developed.The system basically realizes the basic parameter information of nuclear explosion and meteorological parameter information input,and then quickly and conveniently outputs the prediction results of radioactive contamination of nuclear explosion,and then combines the real-time monitoring data and outputs the modified prediction model results of radioactive contamination of nuclear explosion through the optimization and calibration of data assimilation algorithm.The system has a relatively friendly operation interface and rich optional functions,which provides strong support for the emergency management model after nuclear explosion.After theoretical analysis,simulation and data validation,the data assimilation algorithm proposed in this paper can be considered as basically feasible to be applied in the field of nuclear explosion prediction in combination with measured data.The computational prediction results of this method can be used for the prediction of radioactive contamination and hazard trend analysis under nuclear strike or other nuclear weapon explosion scenarios,thus providing decision support for nuclear emergency response and theoretical guidance for post-disaster relief of nuclear strike.