Trend prediction method based on double hidden layer quantum circuit recurrent unit neural network

Abstract

The invention relates to a trend prediction method based on a double hidden layer quantum circuit recurrent unit neural network(DHL-QCRUNN). The trend prediction method comprises the steps of: constructing a permutation entropy set of original operation data; inputting the permutation entropy set into DHL-QCRUNN for training and prediction to obtain a predicted permutation entropy set; constructing a permutation entropy error set of predicted values and actual values at each time point; inputting the permutation entropy error set into DHL-QCRUNN for training and prediction to obtain a predicted and normalized permutation entropy error set; and performing denormalization processing on the predicted and normalized permutation entropy error set to obtain a final prediction result. The invention proposes a novel quantum neural network: the double hidden layer quantum circuit recurrent unit neural network. The trend prediction method updates network parameters of the DHL-QCRUNN by adoptingan LM algorithm to improve the convergence performance of the neural network, compared with other artificial intelligence methods, the DHL-QCRUNN has better nonlinear approximation ability, generalization property and faster convergence speed, the trend prediction method is used for predicting a running trend of a monitored object, and the trend prediction achieves high prediction precision, prediction stability and calculation efficiency.

Description

technical field [0001] The invention relates to the technical field of neural networks, in particular to a method for trend prediction based on double-hidden-layer quantum circuit recurrent unit neural networks. Background technique [0002] Rotating machinery (such as engines, steam turbines, etc.), as key equipment widely used in petrochemical, electric power, metallurgy, coal, nuclear energy, etc. Serious accidents that lead to plane crashes and fatalities. Applying advanced fault diagnosis technology to rotating machinery can play a key role in ensuring the safe operation of equipment, saving maintenance costs and preventing environmental pollution, and has huge economic benefits. [0003] Generally, the maintenance methods of mechanical equipment can be divided into three types: accident shutdown maintenance, regular shutdown maintenance, and condition-based maintenance (also known as predictive maintenance). Among them, condition-based maintenance has good developmen...

AI Technical Summary

Problems solved by technology

However, these prediction methods are difficult to obtain ideal prediction results due to their respective defects

For example, the support vector machine kernel function and its parameters are artificially selected in many cases, with many uncertainties; fuzzy logic has similar problems to SVM; in prediction methods based on artificial neural networks, such as BP neural network (Back-Propagation Neural Network) Network, BPNN), recurrent neural network (Recurrent Neural Network, RNN) and other classic neural networks, there are problems such as slow learning convergence speed, difficult training, and unstable learning and memory of the network, while Elman neural network (Elman Neural Network, Elman -NN) and RNN variants such as Long Short Term Memory Neural Network (LSTMNN), due to their own theoretical and structural defects, it is still difficult to make accurate predictions

Images