Earticle

최근 우리나라에는 기상변화로 인한 영향으로 홍수가 빈번히 발생하고 있다 집중호우는 홍수로 인한 피해를 가중시키는 역할을 하며 매년 많은 사상자와 재산피해를 유발한다. 홍수위험 저감대책에는 구조적인 대책과 비구조적인 대책이 있다. 대부분의 홍수재해 예측의 문제는 비구조적인 대책에 속한다. 신경망 모형은 입력과 출력만을 고려하여 모형을 구성할 수 있기 때문에 비구조적인 문제를 다루기에 적합하다. 인공지능 모형인 신경망 모형을 이용하여 수위예측이 가능한 모형을 구성하고 IHP대표유역 중 하나인 금강 보청천 유역의 기대지점에 적용하였다. 그 결과 신경망 모형은 중소하천유역인 보청천유역에서 홍수위 예측을 위한 우수한 모형으로 판단되었다.

Due to recent unusual climate change, flood happen frequently in Korea. Heavy rainfall increase the damage caused by the flooding. It is cause heavy losses of both life and property every year. Flood hazard mitigation measures consist of structural and non-structural mitigation. Most of flood disaster predictions belong to non-structural mitigation. Neural network is proper to solve non-structural problem. Because it is consider only inputs and outputs to construct model. Real-time water level forecasting model was used to construct artificial intelligence neural network. and it was applied to be a highly suitable tool producing a high water level stage forecasting accuracy at Gidae(No.2) of Bocheong stream, which is IHP representative basins. As a result, neural network was proved to be outstanding model for the water level forecasting in the Bocheong stream catchment.

As a new entertainment and social way, online games now have a huge and increasing user group, so it is of great significance to identify the data stream of online games. Using the excellent nonlinear fitting ability of BP neural network and the advantages of global search of genetic algorithm, the initial weights and thresholds of BP neural network are optimized, and the BP neural network model optimized by genetic algorithm is established. The muti-dimensional input information is proposed to identify online game data streams. Through the experimental simulation, it shows that the selected muti-dimensional information and the established model can be well applied to online game stream recognition.

A strategy which using the particle swarm algorithm improved by the artificial fish swarm algorithm to optimize the BP (Back propagation) algorithm was proposed. It can conquer the shortcomings that the convergence rate of BP is too slow and it is easy to fall into local extreme value, and can improve the learning ability of BP neural network. Finally, the improved algorithm has been used to analysis the earthquake prediction. The results of simulation and test show that the optimized algorithm can improve the predicting accuracy of the BP network.

Based on the theory of BP neural network, we use AHP method to construct the index system and standard of enterprise collaborative management of intellectual property rights under the view of innovation, and the specific evaluation indexes are given from five aspects. In this paper, in order to overcome the existing shortcomings of multi index system In the evaluation method, we use AHP-GA-BP neural network to comprehensive evaluation each index of enterprise intellectual property management. Our method not only displays the evaluation expert experience in learning The evaluation results show that this algorithm can effectively analyze enterprise intellectual property management problems.

Colleges and universities should make full use of modern information technology to improve the teaching mode based on computer aided English teaching mode. The new teaching model should be based on modern information technology, especially network technology, so that English teaching is not subject to the restrictions of time and place, and toward the development of personalized, autonomous learning. In this paper, we analyze network topology structure and put forward an improved RTRL learning algorithm. Based on the empirical analysis, the result shows that with the increase of teaching time, the result of the experimental class is more obvious, so that the new teaching model promotes the improvement of students' autonomous learning ability. In conclusion, cultivating students' autonomous learning ability in a new and diversified foreign language teaching environment has become the core of College English teaching reform.

With the development of network technology, the data capacity become more abundant. How to effectively manage the data, the retrieval more quickly, accurately, improve the data classification accuracy becomes crucial. The BP neural network algorithm with its learning speed, strong ability to adapt and is widely used in network in data mining. Exist but its convergence rate is not high and big error and other shortcomings, therefore, on the basis of traditional algorithm, an improved BP neural network algorithm is put forward. Low error, through experimental analysis, the improved algorithm convergence rate is better.

The efficiency, quality and accuracy of facial image recognition are restricted by luminance, posture, image quality, massive data and method of image recognition, etc. In response to this, this thesis proposes a facial image recognition algorithm based on BP neural network. It improves on traditional BP neutral network by constructing neutrons of facial image recognition in the input layer, hidden layer and output layer. And by constructing the network framework structure of facial image recognition, it also constructs design elements of facial image recognition from input code and output code and therefore constructs the facial image recognition algorithm based on BP neural network. This thesis verifies the algorithm through practical cases and proves that the algorithm is effective and operable.

With the rapid development of Internet technology, the network now has a large size and high complexity, and consequently the network management is becoming increasing difficult and complexity, so traffic forecast play a more and more role in network management. With a large amount of real traffic data collected from the actual network, an adaptive network traffic prediction algorithm based on BP neural network was proposed in this paper, it use an adaptive learning rate method to adjust the learning rate according to total error changing trend of decreased or increased and the difference of changing; and then it corrects the weights in each layers according to forward and reverse calculation. Simulation results show that, compared with the traditional BP neural network, our algorithm has better performance in the prediction results, and has smaller error.

In recent years, the problem of network security has been more and more people's attention, as one of the most important technology of network security, intrusion detection technology has gone through nearly thirty years of development, but it still exists some deficiency factors. Aiming at the defects of the traditional BP neural network intrusion detection model in the detection rate and the convergence speed, the improved PSO-BP neural network is applied to intrusion detection system model in this paper. Experimental and simulation, verifying the improved effect of system in the false negative rate, false positives rate and convergence speed of. Detailed analysis of the standard BP neural network algorithm and improved way of common, including gradient descent algorithm and additional momentum algorithm. Local search capability of BP neural network and the global search ability of particle swarm optimization , we have a detailed description of the PSO algorithm is applied to the case of BP neural network and discusses the improved PSO-BP neural network algorithm flow.

Since the idea of the supply chain management is proposed, many enterprises have attached great importance to the supply chain management and pay a lot of manpower and resources to study. It is also the focus to study the inventory in the field of the supply chain. Quantity of the inventory is not only related to the profit of the enterprises, but also related to the survival of the entire supply chain. Predicting the inventory can improve the ability of enterprises to prevent risk, increase the profits and reduce the losses. In order to predict better on inventory, we propose an improved BP neural network algorithm. In the algorithm, we use the improved GSA algorithm to optimize the parameters of BP neural network algorithm and improve the BP neural network algorithm aiming at its deficiency. The experimental results show that this method has good prediction effect.

The coal and gas outburst is one of complex geological disasters and its prediction is influenced by a multiple of factors, such as coal gas, ground stress, physical and mechanical properties, and complex non-linear system, which cause the low prediction accuracy. It is a favorable scheme to use the nonlinear BP neural network for the prediction algorithm design. But, the traditional BP neural network algorithm has some defects, such as the slow convergence speed and falling into the local minimum value easily. In order to remedy the defects and improve the prediction accuracy of the coal and gas outburst effectively, the improved BP neural network prediction algorithm of the coal and gas outburst is put forward in this paper. The additional momentum is adopted to adjust the network weight and to speed up the network convergence speed, and then the speed of network learning is adjusted self-adaptively and the number of iterations is reduced. Finally, the simulation of prediction of the coal and gas outburst in mine is carried out. Compared with the traditional BP neural network, the improved algorithm shows its superiorities and provides the basis for the accurate prediction of coal mine disasters.

The normal running of the aero-engine is an important guarantee to the aviation aircraft flying in safe. As a result, the analysis and processing of the aero-engine vibration signal is an important task, which can realize the running state monitoring and fault diagnosis to the aviation aircraft. Due to the complexity of the aero-engine’s structure, the vibration signals of aero-engine from the sensors fixed upon the aero-engine’s brake often consist of several signals in aliasing, and also contain noise and other disturbance signal among them. The traditional vibration signal processing methods aiming at the anti-interference and de-noising have no significant effect. In addition, due to the non-linearity of the mixed signal, signal feature recognition and extraction is difficulties. This paper presented an application of the BP neural network to the aero-engine vibration signal separation, through the simulation of the aero-engine vibration signal induced by the high pressure rotor and low pressure rotor rotational imbalance; we proved the accuracy of the algorithm, which can separate the aero-engine vibration signal effectively. Through comparing with the fault spectrum characteristics of the aero-engine, the method can predict and diagnosis the aero-engine fault, which has a very important practical value.

The successful application of short electric arc machining (SEAM) technology can solve the long-standing technical problem of hard-facing materials processing which machinery manufacturing industry generally faces. In the article, we trained high power supply neural network model by using the simulation data. And on this basis, we optimize parameters of power supply combined with genetic algorithm based on objective weighting method to guide parameters changes to meet the requirements. The results, by analyzing startup test and load-mutation test, show that the power supply have the advantages of stable output voltage and fast response speed, which meet the expectant targets and machining requirements. At last, through cutting experiment of SEAM on nickel-based superalloy, the power supply of SEAM designed by this new method is verified that its electric properties meet processing requirements of SEAM.

In the modern supply chain management, optimizing supply chain can reduce the cost of the enterprise. In the optimization of supply chain, the inventory optimization is a very important part. Through forecasting the inventory amount, we can reduce the cost of the inventory. And we also can optimize the supply chain. Because the supply chain network is complex, we use the chaos theory to study and predict the inventory. In this paper, we put forward a chaotic forecasting method which is based on the BP neural and genetic algorithm. This new method is BP-GA chaos prediction algorithm. We use the method to predict the inventory amount. The experiment shows that the method has achieved good forecasting effect.

In this paper, we propose the modified uniformly most powerful (UMP) belief-propagation (BP)-based decoding algorithm which utilizes multiplicative and additive factors to diminish the errors introduced by the approximation of the soft values given by a previously proposed UMP BP-based algorithm. This modified UMP BP-based algorithm shows better performance than that of the normalized UMP BP-based algorithm, i.e., it has an error performance closer to BP than that of the normalized UMP BP-based algorithm on the additive white Gaussian noise channel for low density parity check codes. Also, this algorithm has the same complexity in its implementation as the normalized UMP BP-based algorithm.

In order to solve the undetected probability of multiple targets in ultra-wideband (UWB) through-the-wall radar imaging (TWRI), a time-delay and amplitude modified back projection (BP) algorithm is proposed. The refraction point is found by Fermat's principle in the presence of a wall, and the time-delay is correctly compensated. On this basis, transmission loss of the electromagnetic wave, the absorption loss of the refraction wave, and the diffusion loss of the spherical wave are analyzed in detail. Amplitude compensation is deduced and tested on a model with a single-layer wall. The simulating results by finite difference time domain (FDTD) show that it is effective in increasing the scattering intensity of the targets behind the wall. Compensation for the diffusion loss in the spherical wave also plays a main role. Additionally, the two-layer wall model is simulated. Then, the calculating time and the imaging quality are compared between a single-layer wall model and a two-layer wall model. The results illustrate the performance of the time-delay and amplitude-modified BP algorithm with multiple targets and multiple-layer walls of UWB TWRI.