BP Network Training Research Articles

Inverse tracing of fire source in a single room based on CFD simulation and deep learning

Inverse tracing of fire source location is important for investigation after the fire accidents. In this work, the CFD simulations and deep learning were combined to explore a more efficient and intelligent tool for fire investigation. Firstly, a CFD model for single room was built using FDS. Then abundant simulations were performed by varying the initial conditions to collect massive data including temperature distribution and smoke layer heights in the room. The dataset was divided into two parts for the training and validation of BP network, respectively. The inverse model and forward model with the same data set and neural network parameters were established, which saved the time in adjusting models and producing data and improves the efficiency of research. The shared data sets and neural network parameters did not affect the final prediction results. The accuracy of forward and inverse models is still excellent, reaching a high accuracy. In particular, the accuracy of the inverse model had been improved to more than 99% compared with previous studies. The accuracy of the forward model is more than 80%. Besides, the inverse model's robustness was also examined, the model is still valid when some input features are lost.

Research on the Application of Artificial Intelligence Wireless Network Technology in the Optimization of University Resources

With the increase of teaching resources in universities, allocation for teaching resources has gradually attracted people’s attention. How to optimize and reorganize the numerous teaching resources so as to maximize the teaching efficiency and achieve the best teaching effect is a question that the majority of college management workers must think deeply. Due to the development of computer science, artificial neural networks and wireless networks have made considerable progress. This article optimizes the resources of colleges and universities from these two aspects. First, this work proposes an improved artificial neural network to optimize school resources. The algorithm pointed out that the adaptive genetic algorithm itself has problems such as slow evolution speed and poor individual diversity of the population. After that an improved adaptive genetic algorithm is proposed to optimize BP network training process. By introducing individual evolutionary trend adjustment parameters, a new calculation formula for crossover probability as well as mutation probability is constructed. Improved BP algorithm can effectively avoid the shortcomings of local minima. Second, additional momentum and adaptive learning rate strategies are used to optimize BP network to promote the efficiency for network training. Third, the artificial intelligence network is deployed in the wireless network, and the wireless network will complete the work of data collection, processing, and transmission, so as to optimize the resources of colleges and universities.

A Novel Method of BP Neural Network Based Green Building Design—The Case of Hotel Buildings in Hot Summer and Cold Winter Region of China

With the advent of the big data era, architectural design gradually tends to become more quantified and intelligent. This study proposes a novel green design method for energy-saving buildings based on a BP neural network. This study changed the traditional trial–error mode by evaluating energy consumption based on design performance parameters such as building shape, space, and interface. Instead, energy consumption quota values obtained from statistical data, as well as thermal parameters and energy system parameters in energy-saving standards, were taken as input parameters, and then the design scheme of building shape can be obtained through BP neural network technology. Based on data of 61 hotel buildings in a representative city among a hot summer and cold winter climate zone, the BP neural network model is established to control the building design variables, with 41 kgce/m2·a as its energy-saving design target. Through the energy consumption quota, the trained BP network is applied to predict the optimal architectural design parameters, including the building orientation angle, shape coefficient, window–wall ratio, etc., for twelve building typologies in an area range of 5000~60,000 m2. With recommended control thresholds of quantifiable architectural design elements obtained, this research can provide effective design decision-making suggestions for architects.

Fuzzy model based on local injection of MTX combined with traditional Chinese medicine guided by ultrasound intervention to treat CSP

In view of pregnancy at the uterine scar after cesarean section, there is an urgent need to find a conservative treatment mode with drugs that can improve the success rate of treatment, shorten the conservative treatment time, and cause little harm to the body. This research combines ultrasound intervention technology to construct a fuzzy model to comprehensively evaluate the clinical effects of traditional Chinese medicine combined with western medicine in the treatment of CSP. Moreover, this research combines the current application algorithms of machine vision to explain the relevant principles of ultrasound interventional vision. In addition, this research uses the background subtraction method to extract the target, and uses the machine vision algorithm to extract the ultrasonic detection target object. Finally, this research adjusts the control parameters according to the fuzzy control, and outputs the optimal control plan with BP network training. The results of this study show that the combination of Chinese and Western medicines in the treatment of uterine scars after cesarean section has the advantages of small gestational sacs, short pregnancy, and low blood β-HCG, and it has the advantages of less pain, economy, fewer side effects, and successfully preserves the uterus. Therefore, it is worthy of clinical application.

Application of Back Propagation Neural Network Model in Prediction and Diagnosis of Osteoporosis

The classification model of osteoporosis was established, the diagnostic index system of osteoporosis was analyzed, and the BP neural network system was designed. The data of 1261 people from 402 nuclear families were collected, including age, gender and medical history. The actual data were sampled and tested by using the computational function of ANN, so that the model error could be controlled within a set range. The results of neural network analysis suggested that among the 12 factors influencing bone mineral density, gender had the most obvious effect, followed by height, weight and age. The polymorphism of estrogen receptor PvuII had the greatest impact on bone mineral density (ER PvuII) of all candidate genes, and the factor that had the most serious effect was BGP gene. A large number of BP network training test results were within the expected error range, and the same conclusion was drawn using the same sample and traditional statistical methods.Therisk factors application has certain feasibility and accuracy.

Non-Mechanism Model for Superheater Pollution Diagnosis of Waste Incinerator Based on BP Neural Network

Dust accumulation of boiler heating surface is one of the common problems in power plants, which will affect the unit’s economic and safe operation in serious cases. In this paper, the factors that affect the ash deposition in the superheater of the garbage incinerator are analyzed, and a non-mechanism model for diagnosing the ash deposition pollution in the superheater of the garbage incinerator is established by using 3-layer BP neural network. Parameters such as garbage disposal amount, furnace temperature, primary air flow amount and secondary air flow amount are selected as input vectors, parameters such as secondary superheater inlet smoke temperature and primary superheater inlet temperature are taken as output values, and real-time data of units collected by a power plant DCS system are used as sample sets to train the network after normalization processing. The trained BP network has good effect in superheater pollution diagnosis.

Energy Consumption Prediction of Fused Deposition 3D Printer Based on Improved Regularized BP Neural Network

An energy consumption prediction method based on process parameters and neural network was proposed to study the inherent energy consumption characteristics of open source melt deposition 3D printer and improve energy efficiency. An improved regularized network is used for optimization to avoid over-fitting and under-fitting problems. The orthogonal test sample data were trained in MATLAB environment, and the predictive model between process parameters and energy consumption of open source melt deposition 3D printing was established. The energy consumption prediction results of BP networks and regularized networks are compared by analyzing convergence curves and network training charts. The results show that the BP network training has experienced over-fitting, resulting in a prediction energy consumption error of about 10%. The improved regularization method effectively avoids the over-fitting phenomenon and the error of energy consumption prediction is about 1%. It can effectively improve the calculation efficiency of energy consumption prediction, which shows the accuracy of this method in energy consumption prediction.

Cloud manufacturing service evaluation based on modular neural network

In order to solve the evaluation problem of cloud manufacturing service and then minimise the time of choosing the best service, a comprehensive evaluation index system of cloud manufacturing services is constructed in the light of the characteristics of cloud manufacturing service. The service attributes, manufacturing attributes and network attributes are taken into account from two views of the online characteristics and the offline characteristics of cloud manufacturing service in the evaluation index system. And on this basis, a quantitative evaluation model based on modular neural network (MNN) is proposed. Firstly, the clustering algorithm subtractive-k-means is introduced and used to divide the task into several subtasks based on the idea of modularisation. Then, the subtasks are processed through the trained BP network. Finally, the self-adaptive genetic algorithm is used to optimise the integration weights of each subtask processing module. Combining with the cloud manufacturing evaluation index system, this modular neural network model can be used to effectively evaluate the cloud manufacturing service and help customers quickly select the best cloud manufacturing service.

The Fault Diagnosis of Electric Power Metering System Based on Momentum BP Neural Network

Momentum BP neural network is used to diagnose the fault of high voltage electric power metering system according to the advantage of strong self-adaptation. First of all, the information of 7 operation parameters of metering system is extracted as input values of the neural network, secondly, we collect samples to train and test BP network and use 4-bit binary number to represent the normal circumstance and 7 main fault types, The results show that the momentum BP network model is effective in the fault diagnosis of high voltage electric power metering system.

Design of a Preventive Clearing Blocked System Using PLC Control

Film coating machine by way of spraying the coating solution onto the drug is in a continuous rolling surface. The tablet tends to produce tape smaller even if the spray flow blockage. Then the tablet will crack due to high temperature. Therefore, pharmaceutical tablet coating spray flow rate is a key factor affecting the quality. This article first proposes preventive clearing blocked in the pipeline but not blocking the upcoming clogged. Air purge starts immediately. PLC real-time acquisition is carried out by the pressure at both ends of the pipeline, the pipeline flow liquid contents, when the pressure is detected. The flow reaches a critical point obtained by BP network training, immediately starting the high-pressure air purge. The value of real-time detection uses PLC gun pipeline pressure, flow, etc. It combined weights of the neural element BP neural network model that was trained by the PLC programming. Gun pipe blockage and promptly started high pressure alarm air purge, preventive clearing blocked. Experiments show that preventive clearing blocked pipe blockage can promptly remove the gun.

BP Network Optimization Based on Improved Genetic Algorithm

In view of the virtue and shortage of genetic algorithm and BP network, this paper proposes a new BP network training method based on improved genetic algorithm (IGA-BP). This algorithm uses hierarchical code, adaptive crossover and mutation, pruning similar chromosomes, dynamic supply new chromosomes and other operations, so the network structure and weight are optimized at the same time and the "premature" phenomenon is avoided. The simulation results show that the IGA-BP network architecture is simple, the convergence rate is quick, and has good approximation and generalization ability.

Study of BP Network for a Cylinder Shell’s Support Identification

Some key technologies of neural network for structural damage identification, such as structural design of BP network, input parameter forms, training methods, learning rules and so on, are discussed. Modal parameters of a cylindrical shell structure under different support conditions are tested. Then these data are used as training samples for BP network, and depending on whether using random error, two different BP network is trained. Performance of these trained BP network in different case of test error is studied. The results show that adding a random error to the training samples can effectively improve the recognition ability of a BP network.

Regression Analysis of Ship Principal Dimensions Based on Improved PSO-BP Algorithm

The principal dimensions of naval ships regression analysis via trained BP network by improved PSO is proposed in this paper. Firstly, learning factor is adjusted dynamically, and the improved PSO is implanted in the BP network. Then improved PSO-BP is imported when establishing the regression model of ships’ principal dimensions and comparing its results with the results of polynomial regression. The result shows that BP network trained by improved PSO has higher accuracy and fine character of subsection smooth. Therefore, the model has guidance effect of importance to ship’s top demonstration and preliminary design.

The Design of High-Precision Pressure Transmitter Based on MSP430 Microcontroller

We introduce the high smart pressure sensor’s hardware-circuit design based on Ultra-Low Power MSP430 MCU, the BP network algorithm in temperature compensation because of the pressure sensor’s temperature drift is analyzed, and the collected values of MPM280’s temperature and pressure signals with MCU are used in 4-layer BP network for offline training. We express the suited network module by C language program, and analyze the software-implemented method. The result shows that the MSP430 MCU can reduce system power and its rich exterior equipments can reduce the size of pressure sensor, the training of BP network by use of digital A/D value of MCU conversion reduces the error, the pressure sensor’s precision is improved by BP network, the precision is 0.1%.

Structure Optimization of Slip by the Combination of Artificial Neural Network and Genetic Algorithm

The bridge plug is a staple tool used in downhole operation and the performance of the slips has a directly influence on the oil well productivity and production safety. We raised an optimize method based on BP network and genetic algorithm to make sure the slips satisfy the high temperature and high pressure demands. Establishing the slips system and making finite element analysis by ANSYS, abtaining sixteen group datas to constitute the BP network training samples, establishing the BP simulation model reflecting curvature radius of the slip fluke, dip angle of the fluke, angle of the fluke and distance between flukes using nonlinearity mapping ability of the neural network, applying optimize design for the simulation model using global optimization ability of the genetic algorithm and abtaining the optimum structure parameters of the slip. The optimized results indicate the whole performance of the slips system has increased notably.

Parameter prediction in laser bending of aluminum alloy sheet

Based on the basic platform of BP neural networks, a BP network model is established to predict the bending angle in the laser bending process of an aluminum alloy sheet (1–2 mm in thickness) and to optimize laser bending parameters for bending control. The sample experimental data is used to train the BP network. The nonlinear regularities of sample data are fitted through the trained BP network; the predicted results include laser bending angles and parameters. Experimental results indicate that the prediction allowance is controlled less than 5%–8% and can provide a theoretical and experimental basis for industry purpose.

Generalized Relief Error Networks and Patterns with Lower Errors

For the standard Back-Propagation training algorithm, the objective function is defined as the squared error between the actual and desired output. In [3], we introduced a modular system for training BP-networks by means of the so-called Generalized Relief Error Networks (GREN-networks) approximating the unknown objective function. In comparison with the standard Back-Propagation training algorithm, the proposed system does not require the knowledge of the desired output activity for the proper weight adjustment. During training, the necessary error terms are determined by the pre-trained GREN-network. Furthermore, GREN-networks can be applied for searching patterns similar to those presented to and recalled by the trained BP-network, but with a lower error at the output of the GREN-network. This searching process is performed iteratively by the applied GREN-network. Here, we will focus on analyzing the performance of the above model, especially with regard to its knowledge extraction ability. Experimen...