151 results about "Approximate entropy" patented technology

The invention discloses a short-term load prediction method of a variational mode decomposition and deep belief network. The method comprises the following steps of 1) using a variational mode decomposition method to decompose original historical load data into a series of mode functions with different characteristics; 2) using an approximate entropy to calculate each modal function complexity, merging the modal functions whose approximate entropy values are similar into a new component, and carrying out characteristic analysis on each component; 3) in order to calculating correlation of an influence factor and an output variable, carrying out normalization processing on data; 4) combining a period characteristic of a load, and using a mutual information theory to select an input variable set from aspects of a historical load, a meteorology factor, a date type and the like; and 5) constructing the short-term load prediction method based on the deep belief network (DBN), and verifying method validity through a load prediction scene before 24h. By using the method, short-term load prediction precision is effectively increased and an electric power system load prediction problem can be well solved.

The invention provides a method for quantitatively evaluating the symptoms of tremor of a patient with Parkinson's disease according to the approximate entropy and the cross approximate entropy, belonging to the fields of health care and pattern recognition. The method is characterized by comprising the following steps of: collecting the data of specified tremor of the thumb, collecting the data of specified tremor of the index finger, and grading the specified tremor of the thumb and the specified tremor of the index finger according to a UPDRS (unified Parkinson's disease rating scale); preprocessing the data of tremor; separating a sample training set from a sample testing sample; calculating the approximate entropy and the cross approximate entropy of the data of tremor; constructing the model of a classifier, and verifying the effectiveness of a method. The regularity and the synchronization of the tremor of the patient with Parkinson's disease are disclosed effectively according to the approximate entropy and the cross approximate entropy, and the symptoms of tremor can be accurately and quantitatively classified according to the tremor amplitude, the tremor frequency and other characteristics of the patient. The method is used for objectively evaluating the symptoms of tremor of the patient with Parkinson's disease, and can be applied to the fields of treatment and rehabilitation assessment of the patient with Parkinson's disease and the like.

The invention discloses a coma degree evaluating method based on multiple indexes of non-linearity and complexity, which is used for monitoring the brain state of a patient in a coma, and can realize continuous coma depth monitoring, and coma phase grading and early warning. Multiple complexity indexes such as complexity, Lyapunov exponent, approximate entropy, related dimensions, and the like are extracted from coma electroencephalographic signals by adopting a nonlinear dynamics analytical method, and a comprehensive coma state exponent is obtained by combining a traditional GOS (Glasgow Outcome Scale) evaluation system and a traditional GCS (Glasgow Coma Scale) evaluation system; the coefficients of the correlation between the parameters are obtained by clinical experiments; a coma state grading database is founded and a parameter fusion coefficient is determined on the basis of clinical effects; and finally, complete conscious state grading evaluation exponents are founded to guide the treatment of a patient and make prognosis.

This invention discloses a kind of generating method of real-time monitored brain potential trend chart by adopting computer technique and non-linear dynamics analysis method. It relates to a kind of generating method of brain potential non-linear trend chart. It comprises the following steps: first, collecting original EEG information from different position of scalp; A / D converting to digital EEG information; according to the non-linear dynamics theory and by adopting the trail of state space in time sequence reconstruction system, within one second, converting the conductive digital EEG information to non-linear index that relating to dimension, complexity and approximate entropy, which reflected the current non-linear dynamics characteristic; drawing the non-linear trend chart by using the non-linear index of time sequence, to reflect the non-linear dynamics characteristic variation condition of entire course. This invention can real-time monitor the anaesthesia effect, reflect the variation of anaesthesia effect accurately and promptly.

The invention proposes a rolling bearing health evaluation method based on local characteristic scale decomposition-approximate entropy (APEn) and manifold distance. First, an original vibration signal is decomposed by LCD into a plurality of intrinsic scale components (ISCs); then, the approximate entropy of each ISC is calculated; and finally, the manifold distance between the approximate entropy of the ISCs and the approximate entropy of normal data is calculated, and the calculated manifold distance is normalized into confidence (CV) to express the health degree of a rolling bearing. The normal operation of rolling bearings is particularly important in the modern industrial complex mechanical system, so that rolling bearing performance evaluation is of great significance in prediction and health assessment of the mechanical system. However, as bearing vibration signals are nonlinear and unsteady, it is particularly difficult to accurately extract the characteristics of bearing vibration signals. Local characteristics of signals can be extracted accurately using the method proposed by the invention. Results show that the method proposed by the invention can be used to evaluate the health degree of rolling bearings effectively.

The invention discloses a numerical control machine tool cutting tool wear state real-time monitoring method, which comprises the following steps: collecting sensor signals during machining of a numerical control machine tool in real time, and carrying out signal preprocessing; dividing the machining process into a plurality of working subintervals according to different machining tools, and extracting signal data during machining of each tool of the numerical control machine tool; calculating approximate entropy of the signal obtained during machining of each tool of the numerical control machine tool; and comparing the calculated approximate entropy of each tool and a preset threshold, and judging the tool wear state. The method has the advantages of needing no measuring equipment and needing no construction of an experiment platform, and is fast in prediction speed and high in accuracy; and meanwhile, the tool does not need to be assembled or disassembled, normal machining of the numerical control machine tool is not affected, and real-time monitoring of the wear state of the numerical control machine tool cutting tool can be realized.

The invention discloses a detection method of a detection device for drainage pipeline blockage faults. The method comprises the steps that the detection device for the drainage pipeline blockage faults is arranged; a computer obtains a signal of a normal / blocked pipeline from the receiving end; acoustic response signals in two working conditions of a drainage pipeline are acquired; LMD decomposition is conducted on the acoustic response signals of the normal pipeline and the blocked pipeline; correlation coefficients of all PF components of the normal pipeline and an original signal and correlation coefficients of all PF components of the blocked pipeline and the original signal are calculated by adopting a Pearson correlation coefficient method, and each component of which the correlation coefficient exceeds 15% is taken as an effective PF component signal; energy entropy indexes, approximate entropy indexes and average acoustic pressure indexes of the effective PF components of the normal pipeline and the blocked pipeline are calculated to serve as a characteristic set; an optimal parameter of the characteristic set is sought through a K-CV method; the steps are repeated, and fault identification is conducted on other sections of the pipeline by adopting an SVM classifier with the trained parameter. According to the detection method, the detection procedure is conveniently conducted, and the accuracy and reliability of a detection result are improved.

A method for monitoring the depth of anesthesia is provided for detecting the conscious state of one being anesthetized in the recovery phase or induction phase of anesthesia course in order to facilitate an anesthesiologist to predict exactly the dosage of an anesthetic required. At first, an original electroencephalogram (EEG) is taken from one being tested. Then, the original electroencephalogram is analyzed by approximate entropy to obtain its approximate entropy value. Next, the approximate entropy value is multiplied by 1000 / 17, and the corrected value is assumed as the predicted value of depth of anesthesia. The predicted value of depth of anesthesia represents degree of the conscious state or the depth of anesthesia for the one being tested. The higher the predicted depth of anesthesia value, the more conscious the one being tested is, i.e., in a shallower depth of anesthesia. On the other hand, the lower the predicted depth of anesthesia value, the less conscious the one being tested is, i.e., in a deeper depth of anesthesia.

The invention discloses an urination sensing detection method and an urination sensing detection device. The method comprises the following steps of: acquiring original bladder impedance information and an original electrocardiosignal; extracting urination sensing characteristic parameters comprising an electrocardio approximate entropy, an electrocardio high and low frequency energy ratio, an impedance approximate entropy and an impedance high and low frequency energy ratio; establishing a neural network urination sensing detection model, using the urination sensing characteristic parameters as input vectors of a neural network, training the neural network and determining neural network parameters; and detecting the urination sensing degree of tested personnel by utilizing the trained neural network and outputting a detection result. The urination sensing detection method and the urination sensing detection device utilize the electronic technology to implement urination sensing detection and urination awakening on an enuresis patient, are safe and effective and have no any side effects.