WO2011003347A1 - 连续物理量测量装置及方法 - Google Patents

连续物理量测量装置及方法 Download PDF

Info

Publication number
WO2011003347A1
WO2011003347A1 PCT/CN2010/075008 CN2010075008W WO2011003347A1 WO 2011003347 A1 WO2011003347 A1 WO 2011003347A1 CN 2010075008 W CN2010075008 W CN 2010075008W WO 2011003347 A1 WO2011003347 A1 WO 2011003347A1
Authority
WO
WIPO (PCT)
Prior art keywords
steady state
value
transient
sample value
state process
Prior art date
Application number
PCT/CN2010/075008
Other languages
English (en)
French (fr)
Inventor
郝玉山
Original Assignee
保定市三川电气有限责任公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 保定市三川电气有限责任公司 filed Critical 保定市三川电气有限责任公司
Priority to US13/382,918 priority Critical patent/US8954294B2/en
Priority to CA2767586A priority patent/CA2767586C/en
Priority to EP10796724.2A priority patent/EP2453213B1/en
Publication of WO2011003347A1 publication Critical patent/WO2011003347A1/zh

Links

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01DMEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
    • G01D1/00Measuring arrangements giving results other than momentary value of variable, of general application
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R19/00Arrangements for measuring currents or voltages or for indicating presence or sign thereof
    • G01R19/25Arrangements for measuring currents or voltages or for indicating presence or sign thereof using digital measurement techniques
    • G01R19/2506Arrangements for conditioning or analysing measured signals, e.g. for indicating peak values ; Details concerning sampling, digitizing or waveform capturing
    • G01R19/2509Details concerning sampling, digitizing or waveform capturing

Definitions

  • the present invention relates to continuous physical quantity measurement and recording techniques, and more particularly to digital measurement and recording of temperature, pressure, flow, voltage, current, power, phase angle, etc. in thermal and electrical systems. Background technique
  • the value of the continuous physical quantity remains substantially constant or does not change much for a long period of time, called the steady state process; the short-term transition between the two steady state processes is called the transient process.
  • the sample interval is less than half of the minimum time constant T mm of its transient process. Collecting data at such a sample interval produces massive amounts of data. In the case of telemetry, it takes a lot of communication resources to transmit this data. It is therefore inconvenient to record and transmit.
  • the existing method is to artificially specify a time interval T, which only records or transmits data at multiples of the interval, as shown in Fig. 1.
  • T does not satisfy the theorem, and it must cause aliasing, so that the recorded or transmitted data does not correctly reflect the state and change of the physical process, and the data is wrong and cannot be used.
  • the object of the present invention is to at least solve one of the above problems in the prior art.
  • embodiments of the present invention provide a continuous physical quantity measuring apparatus and method that reduces the amount of data output or transmitted without causing aliasing and accurate data.
  • an embodiment of the present invention provides a continuous physical quantity measuring apparatus including a timing unit for outputting a time base and a time stamp; a sampling unit under which the Nyquist is satisfied to be controlled The time interval of the sample theorem is like a continuous physical quantity, and output ⁇ The sample value x k ; the judgment unit determines whether the sample value x k enters the transient process from the steady state process or enters the steady state process from the transient process; the calculation unit, the ⁇ of the sample unit in the steady state process The sample value calculates a steady state value X; an output unit outputs a start time of the steady state process and the steady state value X according to the judgment result and the time scale, and a start time of the transient process and a state of the transient process The sample value is 3 ⁇ 4.
  • the determining unit performs the determination based on the t distribution of the statistical method or the filter output.
  • a pre-processing unit is further included for performing scale conversion, bad data removal, re-sampling and/or rms calculation on the sample values output by the sample unit.
  • the calculation unit calculates the steady state value according to an averaging algorithm or a low pass filter.
  • the method further includes a recording unit, the start time of the steady state process, the start time of the transient process, and a steady state value corresponding to the start time of the transient process. And the sample values of the transient process are recorded.
  • an embodiment of the present invention provides a continuous physical quantity measuring method, the measuring method comprising the following steps: satisfying a time interval of a Nyquist sample theorem under the control of a timer output time base Sampling the continuous physical quantity, and outputting the sample value x k and the time scale; determining whether the sample value x k enters the transient process from the steady state process or whether it enters the steady state process from the transient process; The time stamp, the start time of the steady state process and the steady state value X obtained by calculating the sample value during the steady state process, and the start time of the transient process and the sample value x k of the transient process.
  • the steady state value X is calculated according to a mean algorithm or a low pass filter.
  • the determining step comprises: calculating a mean and a variance corresponding to the sample value x k ; determining whether (x t - ) / ⁇ obeys the t distribution; if obeying, determining that the sample value x k is not From the steady state process to the transient process or judge the sample value x k from the transient process into the steady state process; if not obey, judge the sample value 3 ⁇ 4 from the steady state process into the transient process or judge the sample value x k not from The transient steady state process enters the steady state process.
  • the t distribution judgment can be based on the following formula:
  • the determining step includes: calculating a mean and a variance corresponding to the sample value x k ; determining whether
  • the determining step comprises: calculating whether the mean value corresponding to the sample value x k satisfies whether
  • the determining step includes: filtering the sample value x k to obtain a filter component of the sample value x k ; respectively determining whether the filter component exceeds a corresponding set upper limit value; And if they are not exceeded, it is judged that the sample value 3 ⁇ 4 has not entered the transient process from the steady state process; if there is a component exceeding, it is judged that the sample value x k enters the transient process from the steady state process.
  • the determining step includes: filtering the sample value x k to obtain a filter component of the sample value x k respectively; and determining whether the filter component exceeds a corresponding set lower limit value; And if there is a component excess, it is judged that the sample value 3 ⁇ 4 has not entered the steady state process from the transient process; if it is not exceeded, it is judged that the sample value x k enters the steady state process from the transient process.
  • a pre-processing step of scaling, bad data culling, re-sampling and/or rms calculation of the sample value xk is further included.
  • the method further includes: a start time of the steady state process, a start time of the transient process, a steady state value corresponding to a start time of the transient process, and a transient process The step of recording the sample value.
  • the invention can greatly reduce the output or transmitted data, avoid the aliasing of the data, improve the accuracy of the data measurement, and ensure that the signal is output or recorded without distortion.
  • Figure 1 is a schematic diagram of the output of the existing continuous physical quantity
  • FIG. 2 is a schematic structural diagram of a continuous physical quantity measuring device according to an embodiment of the present invention.
  • FIG. 3 is a schematic diagram of measurement of data time stamps in a steady state process and a transient process according to an embodiment of the present invention
  • FIG. 4 is a flow chart showing a continuous physical quantity measuring method according to an embodiment of the present invention. detailed description
  • the measuring device includes a timing unit 20, a sampling unit 12, a judging unit 14, a calculating unit 16, and an output unit 18.
  • the timing unit 20 outputs a time base (timed reference signal) and a time scale (generally understood as year, month, day, hour, minute, and second) for timing of the sample unit.
  • the sample unit 12 can sample the continuous physical quantity according to the time interval A t under the output time base control of the timing unit 20, wherein A t satisfies the Nyquist sample theorem.
  • ⁇ ⁇ has a relationship with At, so it is considered to be the above-mentioned variant, and this variant is also common. openly known.
  • the sampling unit 12 outputs the continuously obtained sample value 3 to the judging unit 14, and the judging unit 14 judges whether each of the received sample values x k enters the transient process from the steady state process or enters the steady state from the transient process. State process.
  • the determining unit 14 can determine whether the sample value x k enters the transient process from the steady state process or enters the steady state process from the transient process according to the t distribution of the statistical method. If it is a stationary stochastic process, the probability of falling within 3 ⁇ is within 99.7 % due to the influence of random factors; if X obeys the state distribution, then (X - / obeys t distribution. (Refer to Zhejiang University's Probability Theory and Mathematical Statistics) "Probability Theory and Mathematical Statistics of Higher Education Edition and Festival Poetry", 2nd edition of China Statistics Press. In the measurement, recursive f and, for the kth time, it can be judged whether the data x k is still in steady state. Still does not belong.
  • the judging unit 14 can recursively obtain the corresponding mean value and variance ⁇ for the sample value 3 ⁇ 4 output by the sample unit 12, and thus can judge by judging whether or not x k obeys the t distribution. Whether the sample 3 ⁇ 4 leaves the steady state/transient process enters the transient/steady state process or leaves the transient/steady state process and enters the steady state/transient process.
  • the sample value x k does not enter the transient process from the steady state process or judges that the sample value x k enters the steady state process from the transient process; , judging the value of the sample 3 ⁇ 4 from the steady state process into the transient process or judging that the sample value 3 ⁇ 4 has not entered the steady state process from the transient steady state process.
  • the determining unit 14 may determine, based on the filter, whether the sample value has entered the transient process from the steady state process or whether it has entered the steady state process from the transient process.
  • the filter may be any filter of 1st order or higher (e.g., 2nd order, 3rd order, etc.), such as an nth order Kalman filter or the like.
  • Analyzing unit 14 by filtering preclude sample x k, can be obtained preclude filtered components corresponding sample x k.
  • the number of filtered components obtained is also different.
  • the present invention is not limited to a specific number of filtering components, and any suitable filter may fall within the scope of the present invention.
  • the determining unit 14 respectively determines whether the obtained filtering component exceeds the corresponding set upper limit value, and if not, exceeds, determining that the sample value x k does not enter the transient process from the steady state process, that is, does not leave the steady state process; When the component is exceeded, it is judged that the sample value 3 ⁇ 4 enters the transient process from the steady state process, that is, leaves the steady state process.
  • the filter component exceeds the corresponding set lower limit value, respectively, it can be determined whether the sample value x k enters the steady state process from the transient process. Specifically, if there is a component excess, it is judged that the sample value x k does not enter the steady state process; if none of them exceeds, it is determined that the sample value x k enters the steady state process. Thus, the judging unit 14 judges whether the new data is still in the vicinity of a relatively stable value, or a sudden change in the steady state process occurs and enters the transient process.
  • the judging unit 14 inputs the sample value x k to the calculation unit 16 when it is judged that the sample value has not left the steady state process or it is judged that the sample value has entered the steady state process from the transient state.
  • the calculation unit 16 is configured to perform a steady state value calculation on the sample value of the sample unit 12 during the steady state process.
  • the steady state process can represent its steady state value by a value X, such as an average.
  • the judging unit 14 directly inputs the sample value 3 ⁇ 4 into the output unit 18.
  • the calculating unit 16 can calculate the steady state value of the corresponding output value of the steady state process according to the mean algorithm or the low pass filtering algorithm.
  • the steady-state value X is obtained by the mean algorithm or the low-pass filter algorithm in the steady-state time period to avoid random interference and improve the accuracy of the measurement.
  • the number of values is the steady state value X calculated from these sample values.
  • the set constant is related to the time constant of the physical quantity.
  • the computing unit 16 of the present invention can also obtain a steady state value X by using a weighted mean method or a higher order (e.g., 2nd order, 3rd order, etc.) filter or other mean algorithm.
  • a weighted mean method or a higher order (e.g., 2nd order, 3rd order, etc.) filter or other mean algorithm.
  • the invention is not limited to the specific embodiments described above.
  • the calculation unit 16 uses the new sample value x k to correct the current steady state value X to obtain a new steady state value X corresponding to the steady state process.
  • the X correction is the same as the previous X calculation, and is obtained using the data in the current steady-state process time period, except that the new data x k is further included here.
  • the algorithm of the mean and the variance ⁇ required by the judging unit 14 may be the same as the algorithm for calculating the steady-state value X by the calculating unit 16.
  • the judging unit 14 can output the sample value x k of the sample unit 12 to the calculation unit 16, and the corresponding unit and variance are recursively calculated by the calculation unit 16, and returned to the judging unit 14 according to the dotted arrow shown in the figure.
  • the determining unit 14 determines that the steady state process state change has not occurred according to the value returned by the calculating unit 16 in combination with the specific t distribution, the average value calculated by the calculating unit 16 is taken as The current steady state value X of the steady state process. If it is judged that the steady state process state change occurs in the sample value x k , the average value calculated by the calculation unit 16 is taken as an invalid value.
  • the output unit 18 outputs the start time of the steady state process and its corresponding steady state value X according to the judgment result of the judgment unit 14 and the time scale outputted by the timing unit 20, and the start time and transient state of the output transient process.
  • the sample corresponding value x k .
  • the start time of the transient process represents the end time of the adjacent last steady state process
  • the start time of the steady state process represents the end time of the adjacent previous transient process.
  • the judging unit 14 judges that it is in the steady state process, that is, not in the transient process, and the calculating unit 14 represents a steady state value corresponding to the sampled data in the steady state process by a value X.
  • the sample unit 12 samples at time intervals, a new sample data x k is obtained ; the determination unit 14 determines whether x k is still in the steady state process, and is then corrected by the calculation unit 16 using x k to continue the steady state.
  • the judging unit 14 judges whether it has entered a steady state process for the new data xk ; if it does not enter, the current state is continued, and the judging unit 14 will determine the sample value. 3 ⁇ 4 is output to the output unit 18; otherwise, the start timing of the steady state process (i.e., the end time of the transient process) is output by the output unit 18, and x k is used as the initial steady state data of the steady state process by the computing unit 16. , establishing a new steady state value X corresponding to the steady state process; thus continuing.
  • the algorithm for determining the mean value required by the unit 14 may be the same as the algorithm for calculating the steady-state value X by the calculating unit 16. In this way, the judging unit 14 can directly output the sample value x k of the sample unit 12 to the calculation unit 16, and the corresponding unit and variance are recursively calculated by the calculation unit 16, and returned to the judging unit 14 according to the dotted arrow shown in the figure.
  • the judging unit 14 judges whether the sampling value x k has a steady state process or a transient process change according to the value returned by the calculating unit 16 in combination with the specific t distribution. If the transient process is not entered from the steady state process, the average value calculated by the calculation unit 16 is taken as the current steady state value of the steady state process, and the steady state value is output. X is given to the output unit 18. If it is judged that the steady state process is entered from the transient process, the sample value 3 ⁇ 4 is taken as the initial steady state data of the steady state process, and is output to the calculation unit 16 to establish a new steady state value X corresponding to the steady state process.
  • the threshold value x k is used as the initial transient data of the transient process, and the transient data is provided to the output unit 18 and provided to the computing unit 16 It is used to calculate the mean and variance corresponding to the transient data in the transient process.
  • the calculation unit 16 returns the calculated mean and variance to the judgment unit 14 as a basis for judging the new sample data. If it is determined that the sample value x k has not entered the steady state process from the transient process, the sample value x k is continuously supplied to the calculation unit 16, and the mean and variance of the current transient data are updated and calculated, thereby updating the current judgment unit 14 The basis of the t distribution is judged.
  • the steady state value X corresponding to the end of the steady state process may be output, or the steady state value X corresponding to the different time in the steady state process may be output according to the remote demand.
  • the measuring device of the present invention further includes a recordable unit (not shown) for recording the start time of each steady state process (corresponding to the end time of the transient process), the start of the transient process The time (corresponding to the end time of the steady-state process) is recorded and recorded before the start of the new transient process, ie the steady-state value of its adjacent last steady-state process, and the sample value within the transient process.
  • ⁇ ⁇ 1 ⁇ , ⁇ ⁇ 2 ⁇ , ... are the steady-state values corresponding to the respective steady-state processes.
  • the steady-state value of each steady-state process output may be a set of steady-state values corresponding to different times.
  • T0(t 0)
  • Tl, ⁇ 2, ... are a series of time scales indicating the start time of the steady state process and the start time of the transient process
  • ⁇ x k ⁇ is a series of X corresponding to each transient process.
  • T0, Xl, Tl, ⁇ x k ⁇ , T2, X2, T3, ⁇ x k ⁇ , - are recorded as ⁇ Dq ⁇ -full state records.
  • XI, X2 are a steady state value corresponding to the end time of each steady state process.
  • the above output and recorded steady state data is less than the amount of data obtained by the prior interval T time output one time data (multiple Ts are in a steady state process) and aliasing errors are avoided. Moreover, they truly reflect the characteristics of the system over this period of time, which can be used as an error input for subsequent system calculations.
  • the steady state time is much longer than the transient time, the output or transmitted data is greatly reduced. In addition, it is ensured that the signal is recorded without distortion. Of course, the original signal can also be reproduced from the recorded data.
  • the above output may not include the data set ⁇ 3 ⁇ 4 ⁇ corresponding to the transient process, and the output data is:
  • the above output may not include the steady state value X corresponding to the steady state process, and becomes:
  • T0, ⁇ x k ⁇ , Tl, T2, ⁇ x k ⁇ , T3 , ... , are recorded as transient outputs.
  • the measuring device may further comprise a pre-processing unit (not shown), and the pre-processing unit may pre-process the sample values output by the sample unit 12 and then output the samples directly or re-sampled to the calculation unit 14.
  • the pre-processing includes scale conversion, and may also include bad data culling, and for sinusoidal quantities, such as AC voltage, AC current, AC power, etc., may also include calculation of converting the sample value into an effective value, according to the actual system. , the valid value is also resampled.
  • the output of the preprocessing unit is the effective value of the resampling.
  • FIG. 4 there is shown a flow diagram of a continuous physical quantity measurement method in accordance with an embodiment of the present invention.
  • the continuous physical quantity is sampled according to the predetermined time interval A t , and the sample value and the corresponding time scale are respectively output (step 102 ), wherein A t satisfies the Nyquist
  • the time scale is output by the timer.
  • the timer will correspondingly give the corresponding time scale when outputting each time base, so the sampling time corresponding to each sample value can be determined according to the time base.
  • the start time and end time of the process and the steady state value obtained by calculating the sample value during steady state
  • step 104 can determine whether the sample value enters the transient process from the steady state process or enters the steady state process from the transient process according to the t distribution of the statistical method.
  • the corresponding mean and variance may be recursively obtained.
  • the sample value x k does not enter the transient process from the steady state process or judges that the sample value x k enters the steady state process from the transient process; , judging the value of the sample 3 ⁇ 4 from the steady state process into the transient process or judging that the sample value 3 ⁇ 4 has not entered the steady state process from the transient steady state process.
  • the t-distribution judgment can be in accordance with the following criterion: 1 - 1 ⁇ «/ 2 - where ⁇ is a given risk factor and the value of ⁇ is around 0.05. If the above formula is satisfied, it means that x k obeys the t distribution, otherwise it means disobedience.
  • the rejection of the t distribution can also be reduced to
  • it can be further simplified to
  • ⁇ ⁇ , and ⁇ ⁇ cron, where ⁇ is a given constant, and is generally between 2% and 10% according to the accuracy requirement, preferably about 5%, ⁇ ⁇ is the rated value corresponding to the measured continuous physical quantity.
  • steady state of the sample value x k can also be judged here by using other statistical methods such as Gaussian distribution, and the present invention is not limited to the above specific embodiment.
  • the filtering method can be used to determine whether the sample value enters or leaves the steady state process.
  • the filtering method may be any filtering algorithm of 1st order or higher (for example, 2nd order, 3rd order, etc.), for example, an nth order Kalman filter or the like.
  • the filter component exceeds the corresponding set lower limit value, respectively, it can be determined whether the sample value x k enters the steady state process from the transient process. Specifically, if there is one If the component is exceeded, it is judged that the sample value x k has not entered the steady state process; if it is not exceeded, it is judged that the sample value x k enters the steady state process. Open steady state / transient process.
  • the sample value Xk is subjected to the steady state value calculation.
  • the steady state process can represent its steady state magnitude with a value X, such as an average.
  • a value X such as an average.
  • the steady state value of the steady state process corresponding output sample value may be calculated according to an average algorithm or a low pass filtering algorithm.
  • the steady-state value X is obtained by the mean algorithm or the low-pass filter algorithm in the steady-state time period to avoid random interference and improve the accuracy of the measurement.
  • the number of values is the steady state value X calculated from these sample values.
  • the required mean value can be calculated in the same manner as the steady state value X.
  • the current steady state value X is corrected by using the new sample value x k to obtain a new steady state value X corresponding to the steady state process.
  • the start time of the steady state process and its corresponding steady state value X is the last steady state process adjacent to it.
  • the start time of the steady state process is the end time of the last transient process adjacent to it.
  • a value X is calculated to represent the steady state value of the corresponding sampled data in the steady state process.
  • a new sample data x k is obtained ; whether x k is still in the steady state process, and then, x is corrected by x k to continue the steady state process; if not, the steady state process is output.
  • the end time i.e., the time at which x k enters the transient process
  • the steady state value X of the steady state process and x k is output as the data of the transient process.
  • the steady-state value output For the steady-state value output, according to the judgment result, only the steady-state value X corresponding to the end of the steady-state process can be output, or the steady-state value X corresponding to the different time in the steady-state process can be output according to the remote demand.
  • the present invention further includes a recordable step of recording a start time of the steady state process (corresponding to the end time of the transient process) and a start time of the transient process (corresponding to the end time of the steady state process), And record the steady state value of the adjacent last steady state process before the start of the new transient process, and the sample value within the transient process.
  • the above output and recorded steady state data is less than the amount of data obtained by outputting the data at the interval T time (multiple Ts are in a steady state process) and aliasing errors are avoided. Moreover, they truly reflect the characteristics of the system over this period of time, which can be used as an error-free input for subsequent system calculations.
  • the steady state time is much longer than the transient time, the output or transmitted data is greatly reduced. In addition, it is ensured that the signal is recorded without distortion. Of course, the original signal can also be reproduced from the recorded data.
  • a pre-processing step may also be included for pre-processing the sampled sample values and then directly or re-sampling the output for subsequent determination and calculation.
  • the pre-processing includes scale conversion, and may also include bad data culling, and for sinusoidal quantities, such as voltage, current, Power, etc., can also include calculation of effective values.
  • the output sample value is a valid value.
  • the measuring apparatus and method of the present invention may not directly output the time stamp of the start and end of the steady state process, that is, the time stamp is ambiguously output but is implicit in the polling.
  • the output data is in the steady state process, only the corresponding steady state value X is output.
  • the invention can be applied to measurement and recording of temperature, pressure, flow, voltage, current, power and phase angle in thermal power and power systems, can truly and clearly reflect changes of system parameters, and significantly reduce recording and output.
  • the amount of data can be applied to measurement and recording of temperature, pressure, flow, voltage, current, power and phase angle in thermal power and power systems, can truly and clearly reflect changes of system parameters, and significantly reduce recording and output. The amount of data.

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Indication And Recording Devices For Special Purposes And Tariff Metering Devices (AREA)
  • Testing Or Calibration Of Command Recording Devices (AREA)
  • Complex Calculations (AREA)
  • Testing And Monitoring For Control Systems (AREA)

Description

连续物理量测量装置及方法
技术领域
本发明涉及连续物理量的测量和记录技术, 尤其是涉及热工、 电力系 统中的温度、 压力、 流量、 电压、 电流、 功率、 相角等的数字测量和记录。 背景技术
连续物理量的值在相当长一段时间维持基本不变或变化不大, 称为稳 态过程; 两个稳态过程之间的短时间的过渡称为暂态过程。 对于稳态过程 和暂态过程组成的过程物理量釆样, 按照奈奎斯特 (也称作香农) 釆样定 理, 釆样间隔要小于其暂态过程的最小时间常数 T mm的一半。 按照这样的 釆样间隔釆集数据会产生海量数据。 在遥测情况下, 需要占用大量通信资 源传送这些数据。 因此不便记录和传送。
为了减少记录和传送的数据量, 现有的办法是人为规定一个釆样时间 间隔 T, 只记录或传送该间隔整倍数点上的数据, 如图 1 所示。 显然, T 不满足釆样定理, 必定造成混迭, 致使记录或传送的数据不能正确反映物 理过程的状态和变化, 数据错误而无法使用。
如果取 T时间段内的测量平均值替代瞬时值, 也不能解决釆集数据的 错误, 除了产生混迭外, 还会产生额外误差, 使本来积等于两数相乘的三 个量, 其平均值不再存在积的关系。 例如, 变量 A、 B、 C , 其中, A = B . C,B = B + Ab, C = + Ac,I≠B . 。 发明内容
本发明的目的旨在至少解决现有技术中的上述问题之一。
为此, 本发明的实施例提出一种既减少输出或传送的数据量, 又不产 生混迭的、 数据 'J量准确的连续物理量测量装置和方法。
根据本发明的一个方面, 本发明的实施例提出了一种连续物理量测量 装置, 包括用来输出时基和时标的定时单元; 釆样单元, 在所述时基控制 下以满足奈奎斯特釆样定理的时间间隔对连续物理量进行釆样, 并输出釆 样值 xk; 判断单元, 判断釆样值 xk是否从稳态过程进入暂态过程或者是否 从暂态过程进入稳态过程; 计算单元, 在稳态过程中对所述釆样单元的釆 样值计算稳态值 X; 输出单元, 根据判断结果以及所述时标, 输出稳态过 程的开始时间和所述稳态值 X , 以及暂态过程的开始时间和所述暂态过程 的釆样值¾。
根据本发明进一步的实施例, 所述判断单元根据统计法的 t 分布或者 滤波器输出进行判断。
根据本发明进一步的实施例, 还包括预处理单元, 用于对所述釆样单 元输出的釆样值进行标度变换、 坏数据去除、 重抽样和 /或有效值计算。
根据本发明进一步的实施例, 所述计算单元根据均值算法或低通滤波 算器计算所述稳态值。
根据本发明进一步的实施例, 还包括记录单元, 所述记录单元对所述 稳态过程的开始时间、 所述暂态过程的开始时间、 在所述暂态过程开始时 间前对应的稳态值以及所述暂态过程的釆样值进行记录。
根据本发明的另一方面, 本发明的实施例提出一种连续物理量测量方 法, 所述测量方法包括以下步骤: 在定时器输出时基的控制下以满足奈奎 斯特釆样定理的时间间隔对连续物理量进行釆样,并输出釆样值 xk和时标; 判断釆样值 xk是否从稳态过程进入暂态过程或者是否从暂态过程进入稳态 过程; 以及根据判断结果以及所述时标, 输出稳态过程的开始时间和在稳 态过程中对釆样值计算获得的稳态值 X , 以及暂态过程的开始时间和所述 暂态过程的釆样值 xk
根据本发明进一步的实施例, 所述稳态值 X根据均值算法或低通滤波 器计算获得。
根据本发明进一步的实施例, 所述判断步骤包括: 计算釆样值 xk对应 的均值 和方差^ ; 判断(xt - )/ ^是否服从 t分布; 若服从, 判断釆样值 xk未从稳态过程进入暂态过程或者判断釆样值 xk从暂态过程进入稳态过 程; 若不服从, 判断釆样值 ¾从稳态过程进入暂态过程或者判断釆样值 xk 未从暂稳态过程进入稳态过程。 所述 t 分布判断可以依据以下公式:
\xk
Figure imgf000004_0001
其中 k为釆样值的数量, "为风险系数, 一般取值 0.05左右。
根据本发明进一步的实施例, 所述判断步骤包括: 计算釆样值 xk对应 的均值 和方差 ; 判断是否满足 |xt - |≤ ^ , 其中 A为给定值, 一般在
3 ~ 10之间; 若满足, 判断釆样值 xk未从稳态过程进入暂态过程或者判断 釆样值 xk从暂态过程进入稳态过程; 若不满足, 判断釆样值 xk从稳态过程 进入暂态过程或者判断釆样值 xk未从暂稳态过程进入稳态过程。
根据本发明进一步的实施例, 所述判断步骤包括: 计算釆样值 xk对应 的均值 判断是否满足 |xt - |≤ χ„, 其中 δ为给定值, 一般在 2%到 10% 之间, χη为所述连续物理量对应的额定值; 若满足, 判断釆样值 xk未从稳 态过程进入暂态过程或者判断釆样值 xk从暂态过程进入稳态过程; 若不满 足,判断釆样值 xk从稳态过程进入暂态过程或者判断釆样值 ¾未从暂稳态 过程进入稳态过程。
根据本发明进一步的实施例, 所述判断步骤包括: 对釆样值 xk进行滤 波, 分别得到釆样值 xk的滤波分量; 分别判断所述滤波分量是否超出对应 设定的上限值; 以及若均未超出, 判断釆样值 ¾未从稳态过程进入暂态过 程; 若存在一个分量超出, 判断釆样值 xk从稳态过程进入暂态过程。
根据本发明进一步的实施例, 所述判断步骤包括: 对釆样值 xk进行滤 波, 分别得到釆样值 xk的滤波分量; 分别判断所述滤波分量是否超出对应 设定的下限值; 以及若存在一个分量超出, 判断釆样值 ¾未从暂态过程进 入稳态过程; 若均未超出, 判断釆样值 xk从暂态过程进入稳态过程。
根据本发明进一步的实施例, 还包括对釆样值 xk进行标度变换、 坏数 据剔除、 重抽样和 /或有效值计算的预处理步骤。
根据本发明进一步的实施例, 还包括对所述稳态过程的开始时间、 所 述暂态过程的开始时间、 在所述暂态过程开始时间前对应的稳态值以及所 述暂态过程的釆样值进行记录的步骤。
本发明可以将输出或传送的数据大幅度减少, 同时可以避免数据出现 混迭, 提高数据测量准确性, 保证了信号不失真地被输出或记录下来。
本发明附加的方面和优点将在下面的描述中部分给出, 部分将从下面 的描述中变得明显, 或通过本发明的实践了解到。 附图说明
本发明的上述和 /或附加的方面和优点从下面结合附图对实施例的描 述中将变得明显和容易理解, 其中:
图 1为现有连续物理量的釆样输出示意图;
图 2为本发明实施例的连续物理量测量装置结构示意图;
图 3为本发明实施例的稳态过程和暂态过程中数据及时标的测量示意 图;
图 4显示了本发明实施例的连续物理量测量方法流程图。 具体实施方式
下面详细描述本发明的实施例, 所述实施例的示例在附图中示出, 其 中自始至终相同或类似的标号表示相同或类似的元件或具有相同或类似功 能的元件。 下面通过参考附图描述的实施例是示例性的, 仅用于解释本发 明, 而不能解释为对本发明的限制。
现在参考图 2 , 该图为本发明实施例的连续物理量测量装置结构示意 图。 如图 2 中所示, 测量装置包括定时单元 20、 釆样单元 12、 判断单元 14、 计算单元 16和输出单元 18。 定时单元 20输出时基(定时的基准信号 ) 和时标 (一般可以理解为年月日时分秒) , 用于釆样单元的定时。 釆样单 元 12可以在定时单元 20的输出时基控制下,按照时间间隔 A t对连续物理 量进行釆样, 其中 A t满足奈奎斯特釆样定理。
为了简化正弦量的计算, 按照等相位 Δ φ釆样 (如锁相环输出控制的 釆样) , Δ φ与 A t存在关系, 所以, 认为是上述釆样的变种, 而这种变种 也是普遍公知的。
釆样单元 12将不断获得的釆样值 ¾输出给判断单元 14 , 判断单元 14 则判断接收到的每个釆样值 xk是否从稳态过程进入暂态过程或者是否从暂 态过程进入稳态过程。
在一个实施例中, 判断单元 14可以根据统计法的 t分布判断釆样值 xk 是否从稳态过程进入暂态过程或者从暂态过程进入稳态过程。 如果为平稳随机过程, 由于随机因素的影响, 落在 3 σ内概率在 99.7 %之内; 如果 X服从状态分布, 则(X - / 服从 t分布。 (参考浙江大学的 《概率论与数理统计》 高教版和節诗书的 《概率论与数理统计》 中国统计 出版社的第 2版) 。 在测量中, 递推 f和 , 对于第 k次, 就能够判断数 据 xk是否仍属于稳态还是不属于。
在本发明的实施例中,判断单元 14对于釆样单元 12输出的釆样值 ¾, 可以递推得到其对应的均值 和方差^ , 因此通过判断 (xk 是否服从 t分布, 从而可以判断釆样值 ¾是否离开稳态 /暂态过程进入暂态 /稳态过程 或者离开暂态 /稳态过程而进入稳态 /暂态过程。
具体来说, 若判断(xt - )/ 服从 t分布, 则釆样值 xk未从稳态过程进 入暂态过程或者判断釆样值 xk从暂态过程进入稳态过程; 若不服从, 判断 釆样值 ¾从稳态过程进入暂态过程或者判断釆样值 ¾未从暂稳态过程进入 稳态过程。
在一个实施例中,判断单元 14可以根据滤波器判断釆样值是否从稳态 过程进入暂态过程或者是否从暂态过程进入稳态过程。 这里, 滤波器可以 是任意 1阶或更高阶(例如 2阶、 3阶等) 的滤波器, 例如 n阶卡尔曼滤 波器等。
判断单元 14通过对釆样值 xk进行滤波, 从而可以得到釆样值 xk对应 的滤波分量。 显然, 对于不同阶的滤波器, 得到的滤波分量个数也不同。 这里, 本发明不局限于具体个数的滤波分量, 任意合适的滤波器均可落在 本发明的保护范围内。
判断单元 14分别判断所得的滤波分量是否超出对应设定的上限值,若 均未超出,判断釆样值 xk未从稳态过程进入暂态过程, 即未离开稳态过程; 若存在一个分量超出, 判断釆样值 ¾从稳态过程进入暂态过程, 即离开稳 态过程。
类似地, 通过分别判断上述滤波分量是否超出对应设定的下限值, 则 可以判断釆样值 xk是否从暂态过程进入稳态过程。 具体来说, 若存在一个 分量超出, 判断釆样值 xk未进入稳态过程; 若均未超出, 判断釆样值 xk 进入稳态过程。 这样, 判断单元 14判断新数据是否仍处于相对稳定的数值附近, 还是 出现了超出稳态过程的突变而进入暂态过程。
在判断釆样值未离开稳态过程或者判断釆样值从暂态进入稳态过程 时, 判断单元 14将釆样值 xk输入给计算单元 16。 计算单元 16用于在稳态 过程中, 对釆样单元 12的釆样值进行稳态值计算。 对于连续物理量, 在稳 态过程能够以一个值 X , 例如平均值代表其稳态值大小。 对于暂态过程的 釆样值, 判断单元 14将釆样值¾直接输入到输出单元 18中。
计算单元 16 可以根据均值算法或低通滤波算法计算稳态过程对应输 出釆样值的稳态值。 稳态值 X通过稳态时间段内的均值算法或低通滤波算 法获得, 以避免随机干扰、 提高测量的精度。
在一个实施例中, 稳态值 X 对应的均值算法公式例如表示为 ^ = 1^ 0 . ^4 + ^ )] , 其中 表示稳态过程获得的釆样值, k表示这些釆样 k
值的个数, 即为根据这些釆样值计算得到的稳态值 X。 k = 1时,则
在一个实施例中,稳态值 X可以根据低通滤波器的递推算法计算获得, 例如利用公式表示为 = «. U (l - «) · χ^々一阶低通滤波算法, 其中 α为设 定的常数, 与物理量的时间常数有关。
当然, 本领域技术人员显然可知, 本发明计算单元 16还可以釆用加权 均值法或更高阶(例如 2 阶、 3 阶等) 滤波器或者其他均值算法来获得稳 态值 X。 本发明不局限于上述具体实施例。
对于上述稳态值 X计算的递推算法,在稳态过程中,计算单元 16利用 新的釆样值 xk对当前稳态值 X进行修正,得到对应稳态过程的新稳态值 X。 这里, 对 X修正与之前 X的计算相同, 利用当前稳态过程时间段内的釆样 数据获得, 只不过这里进一步包括了新数据 xk
在判断单元 14利用统计法的 t分布进行判断的实施例中,判断单元 14 所需的均值 和方差^的算法,可以与计算单元 16计算稳态值 X的算法相 同。这样,判断单元 14可以将釆样单元 12的釆样值 xk输出给计算单元 16 , 由计算单元 16递推计算得到对应的均值和方差,并按图示虚线箭头返给判 断单元 14。 若判断单元 14根据计算单元 16返回的数值结合具体 t分布判 断釆样值 ¾未发生稳态过程状态变化, 则将计算单元 16计算的均值作为 该稳态过程的当前稳态值 X。 若判断釆样值 xk发生稳态过程状态变化, 则 将计算单元 16计算的均值作为无效值。
输出单元 18则根据判断单元 14的判断结果,并结合定时单元 20输出 的时标, 输出稳态过程的开始时间和及其对应的稳态值 X, 以及输出暂态 过程的开始时间和暂态过程对应的釆样值 xk。 当然, 暂态过程的开始时间 即表示其相邻的上一个稳态过程的结束时间,稳态过程的开始时间即表示 其相邻的上一个暂态过程的结束时间。
具体来说, 对于连续物理量, 假定判断单元 14判断其以往位于稳态过 程, 即不在暂态过程, 计算单元 14以一个值 X代表该稳态过程中对应已 釆样数据的稳态值大小。 在釆样单元 12以时间间隔釆样之后, 获得一个 新釆样数据 xk; 判断单元 14判断 xk是否仍在稳态过程, 是, 则由计算单 元 16利用 xk修正 X, 继续稳态过程; 不是, 表示 xk从稳态过程进入暂态 过程, 则由输出单元 18输出该稳态过程的结束时间 (即 ¾进入暂态过程 的开始时间)、 该稳态过程的稳态值 X, 并将 xk作为暂态过程的釆样数据 输出。
另外, 如果以往不在稳态过程, 即处于暂态过程, 对于新数据 xk判断 单元 14判断其是否进入到一个稳态过程; 若没有进入, 则继续当前状态, 由判断单元 14将釆样值 ¾输出给输出单元 18; 否则, 由输出单元 18输 出稳态过程的开始时标 (即该暂态过程的结束时间) , 并由计算单元 16 将 xk作为该稳态过程的初始稳态数据, 建立该稳态过程对应的新的稳态 值 X; 如此继续。
在判断单元 14利用统计法的 t分布进行判断的实施例中,判断单元 14 所需的均值 的算法,可以与计算单元 16计算稳态值 X的算法相同。这样, 判断单元 14可以将釆样单元 12的釆样值 xk直接输出给计算单元 16 , 由计 算单元 16递推计算得到对应的均值和方差,并按图示虚线箭头返给判断单 元 14。
判断单元 14根据计算单元 16返回的数值结合具体 t分布判断釆样值 xk是否发生稳态过程或暂态过程变化。 若未从稳态过程进入暂态过程, 则 将计算单元 16计算的均值作为该稳态过程的当前稳态值, 并输出稳态值 X给输出单元 18。 若判断从暂态过程进入稳态过程, 则以釆样值 ¾作为 该稳态过程的初始稳态数据, 并输出给计算单元 16建立该稳态过程对应 的新的稳态值 X。
若判断釆样值 xk从稳态过程进入暂态过程,则以釆样值 xk作为该暂态 过程的初始暂态数据, 将暂态数据提供给输出单元 18 , 并提供给计算单 元 16用于计算暂态过程中暂态数据对应的均值和方差。计算单元 16将计 算的均值和方差返给判断单元 14作为对新釆样数据的判断依据。 若判断 釆样值 xk未从暂态过程进入稳态过程, 则继续将釆样值 xk提供给计算单 元 16 , 对当前暂态数据的均值和方差进行更新计算, 从而更新判断单元 14当前的 t分布判断依据。
对于输出单元 18, 其可以根据判断单元 14 的判断, 仅输出在稳态过 程结束时对应的稳态值 X, 也可以根据远程需要, 将稳态过程中不同时间 对应的稳态值 X输出。
在一个实施例中, 本发明的测量装置还包括可以记录单元 (图中未显 示),记录单元对对每个稳态过程的开始时间(对应暂态过程的结束时间)、 暂态过程的开始时间 (对应稳态过程的结束时间)进行记录, 并且记录在 新的暂态过程开始时间前, 即其相邻上一个稳态过程的稳态值, 以及该暂 态过程内的釆样值。
例如对于图 3所示釆样的连续物理量, 按照上述方式, 输出单元 18对 其稳态过程和暂态过程可以输出到的数据是:
TO, { Xl },Tl, {xk},T2, { X2},T3, {xk} , ... ,
其中, { Χ1 }、 { Χ2}、 …为各个稳态过程对应的稳态值, 如上文所示, 每个稳态过程输出的稳态值可以是不同时间对应的稳态值集合。 T0( t=0 )、 Tl、 Τ2、 …为表示对应稳态过程开始时间和暂态过程开始时间的一系列时 标, {xk}为每个暂态过程中对应的一系列 X的釆样值。
对于记录单元, 则其可以记录数据如下:
T0,Xl,Tl, {xk},T2,X2,T3, {xk} , - , 记为 {Dq}-全态记录。 这里, XI、 X2 是对应每个稳态过程结束时间的一个稳态值。
对于开始为暂态过程的情况, 则对应输出如下: T0,{xk},Tl, { Xl },T2,{xk},T3, { X2} .., 是稳态过程与暂态过程对应输 出的另一种表现, 与图 3实施例无实质区别。
一般地, 以上输出和记录的稳态数据比现有的间隔 T时间输出一次数 据获得的数据量少(多个 T都在一个稳态过程 )且避免了混迭错误。并且, 它们真实反映了系统在该时间段上的特性,从而可作为后续系统计算的无 误输入。
由于稳态时间较暂态时间要长得多, 所以输出或传送的数据大幅度减 少。 另外, 又保证了信号不失真地被记录下来。 当然, 由记录的数据也能 够复现原信号。
如果只关心稳态值, 即稳态过程测量, 则上述的输出中可以不包括暂 态过程对应的釆样数据集 {¾} , 输出的数据是:
TO, { X1 },T1, Τ2, { Χ2},Τ3, ... , 记为稳态输出。
而如果只关心暂态, 即暂态过程测量, 上述的输出可以不包括稳态过 程对应的稳态值 X, 而变成:
T0,{xk},Tl, T2,{xk}, T3 , ... , 记为暂态输出。
在一个实施例中, 测量装置还可以包括预处理单元 (图中未显示) , 预处理单元可以对釆样单元 12输出的釆样值进行预处理, 然后直接或重抽 样输出到计算单元 14。 一般地, 预处理包括标度变换, 还可以包括坏数据 剔除, 而对于正弦量, 例如交流电压、 交流电流、 交流功率等, 还可以包 括将釆样值转换为有效值的计算, 依据实际系统, 有效值还要重抽样。 在 这种情况下, 预处理单元的输出是重抽样的有效值。
现在参考图 4 , 该图显示了本发明实施例的连续物理量测量方法流程 图。
首先, 可以在定时器输出时基的控制下, 按照预定时间间隔 A t对连续 物理量进行釆样, 并分别输出釆样值和对应的时标 (步骤 102 ) , 其中 A t 满足奈奎斯特釆样定理, 时标由定时器输出。 定时器在输出每个时基时会 相应地给出对应的时标, 因此根据时基可确定每个釆样值对应的釆样时间。 然后, 判断釆样值是否从稳态过程进入暂态过程或者是否从暂态过程进入 稳态过程 (步骤 104 ) , 根据对应的判断结果以及所述时标, 输出稳态过 程的开始时间和结束时间以及在稳态过程中对釆样值计算获得的稳态值
(步骤 106) 。
在一个实施例中, 步骤 104可以根据统计法的 t分布判断釆样值是否 从稳态过程进入暂态过程或者从暂态过程进入稳态过程。
对于统计法的 t 分布: 在方差分析中, 如果为平稳随机过程, 由于随 机因素的影响, 落在 3 σ内概率在 99.7%之内。 现实中, 若 σ未知, 可以 用 代替, 且 (χ- )/ 服从 t分布, 在测量中, 递推 f和 , 对于第 k次, 就能够判断数据 xk是否仍属于稳态还是不属于。
在本发明的实施例中, 对于釆样输出的釆样值 xk, 可以递推得到其对 应的均值 和方差 。
例如, 根据均值公式: = [( -1)· 以及方差公式: - 获得对应的均值和
Figure imgf000012_0001
方差, 其中 k为稳态过程或者暂态过程对应的釆样值的数量。
因此通过判断 (xk - 是否服从 t分布,从而可以判断釆样值 ¾是否 离开稳态 /暂态过程进入暂态 /稳态过程或者离开暂态 /稳态过程而进入稳态 / 暂态过程。
具体来说, 若判断(xt- )/ 服从 t分布, 则釆样值 xk未从稳态过程进 入暂态过程或者判断釆样值 xk从暂态过程进入稳态过程; 若不服从, 判断 釆样值 ¾从稳态过程进入暂态过程或者判断釆样值 ¾未从暂稳态过程进入 稳态过程。
― _― ~ t{k - 1)
如上文所述, 在一个实施例中, t分布判断可以依照以下 判据公式: 1 - 1 «/2 - 其中 α为给定的风险系数, α值在 0.05 左右。 若满足上述公式, 则表示 xk服从 t分布, 否则表示不服从。
1 )对于釆样和预处理输出的 xk;
2 ) 若 k = 1: U'J j = j , =0
否则, 按上述公式计算 和 。
3)若 \xk -xk\≤ fk - ta/2 (k-V)-sk , 则表示为稳、态过程。 即表示 xk未从稳、态 过程进入暂态过程, 即未离开稳态过程或者表示 ¾从暂态过程进入稳态过 程, 即离开暂态过程, 且稳态值; r = 。
对于 xk未离开稳态过程的, ^ k = k+l , 并根据步骤 2和步骤 3继续下 一个釆样值的判断; 对于 xk为进入稳态过程的, 令 k=l , 将¾作为该稳态 过程的初始稳态数据 X1 , 并输出时标, 然后重复上述步骤。
否则, 表示从稳态过程离开而进入暂态过程, 或者未从暂态过程进入 稳态过程。
对于 xk进入暂态过程的, 将 xk作为该暂态过程的初始暂态数据 Xl , 令 k = 1 , 并根据步骤 2和步骤 3继续下一个釆样值的判断; 对于 xk为未离开 暂态过程的, 令1^= 1^+1 , 并重复上述步骤 2和步骤 3。
在一个实施例中, t分布的判拒也可以简化成 |xt - |≤ 其中 A为 给定常数,例如 A位于 3 ~ 10之间。或者, 更进一步可以简化成 |xt - |≤Δ , 而 Δ = χ„, 其中 δ为给定常数, 根据精度要求一般在 2% ~ 10%之间取值, 优选为 5%左右, χη为被测连续物理量对应的额定值。
并且在第 3 ) 步的判断中, 若对于初始测量 (t=0 ) 的釆样值 Xl , 若判 断为稳态过程, 则记录该稳态过程的开始时间, 初始化为 0。
当然, 这里也可以釆用高斯分布等其他统计法对釆样值 xk的稳态状态 进行判断, 本发明不局限于上述具体实施例。
在一个实施例中, 可以根据滤波法判断釆样值是否进入或者离开稳态 过程。 这里, 滤波法可以是任意 1 阶或更高阶(例如 2阶、 3 阶等) 的滤 波算法, 例如 n阶卡尔曼滤波器等。
通过对釆样值 xk进行滤波, 从而可以得到釆样值 xk对应的滤波分量。 显然, 对于不同阶的滤波算法, 得到的滤波分量个数也不同。 这里, 本发 明不局限于具体个数的滤波分量, 任意合适的滤波算法均可落在本发明的 保护范围内。
分别判断所得的滤波分量是否超出对应设定的上限值, 若均未超出, 判断釆样值 xk未从稳态过程进入暂态过程, 即未离开稳态过程; 若存在一 个分量超出, 判断釆样值 xk从稳态过程进入暂态过程, 即离开稳态过程。
类似地, 通过分别判断上述滤波分量是否超出对应设定的下限值, 则 可以判断釆样值 xk是否从暂态过程进入稳态过程。 具体来说, 若存在一个 分量超出, 判断釆样值 xk未进入稳态过程; 若均未超出, 判断釆样值 xk 进入稳态过程。 开稳态 /暂态过程。
这样, 可以判断新数据是否仍处于相对稳定的数值附近, 还是出现了 超出稳态过程的突变而进入到暂态过程。
在判断釆样值未离开稳态过程或者判断釆样值从暂态进入稳态过程 时, 将釆样值 Xk进行稳态值计算。 对于连续物理量, 在稳态过程能够以一 个值 X, 例如平均值代表其稳态大小。 对于判断为暂态过程的釆样值, 则 可以直接输出。
这里, 可以根据均值算法或低通滤波算法计算稳态过程对应输出釆样 值的稳态值。稳态值 X通过稳态时间段内的均值算法或低通滤波算法获得, 以避免随机干扰、 提高测量的精度。
在一个实施例中, 稳态值 X 对应的均值算法公式例如表示为 ^ = 1^ 0. ^4 + ^ )] , 其中 表示稳态过程获得的釆样值, k表示这些釆样 k
值的个数, 即为根据这些釆样值计算得到的稳态值 X。 k = 1时,则
在一个实施例中,稳态值 X可以根据低通滤波器的递推算法计算获得, 例如利用公式表示为 = «. U (l - «) · χ^々一阶低通滤波算法, 其中 α为设 定的常数。
当然, 本领域技术人员显然可知, 本发明还可以釆用更高阶(例如 2 阶、 3阶等)滤波算法或者其他均值算法来获得稳态值 X。 本发明不局限于 上述具体实施例。
在利用统计法的 t 分布进行判断的实施例中, 所需的均值 可以与稳 态值 X的计算方法相同。
对于上述稳态值 X计算的递推算法, 利用新的釆样值 xk对当前稳态值 X进行修正, 得到对应稳态过程的新稳态值 X。
根据判断结果, 并结合定时器输出的时基, 输出稳态过程的开始时间 和及其对应的稳态值 X, 以及输出暂态过程的开始时间和暂态过程对应的 釆样值 xk。 当然, 暂态过程的开始时间即表示其相邻的上一个稳态过程的 结束时间, 稳态过程的开始时间即表示其相邻的上一个暂态过程的结束时 间。
具体来说, 对于连续物理量, 假定判断其以往位于稳态过程, 则计算 一个值 X代表该稳态过程中对应已釆样数据的稳态值大小。 在以时间间隔 釆样之后, 获得一个新釆样数据 xk; 判断 xk是否仍在稳态过程, 是, 则利 用 xk修正 X, 继续稳态过程; 不是, 则输出该稳态过程的结束时间 (即 xk 进入暂态过程的开始时间) 以及该稳态过程的稳态值 X, 并将 xk作为暂态 过程的釆样数据输出。
另外, 如果以往不在稳态过程, 即处于暂态过程, 对于新数据 xk判断 其是否进入到一个稳态过程; 若没有进入, 则继续当前状态, 将釆样值 xk输出; 否则, 输出稳态过程的开始时标 (即该暂态过程的结束时间) , 并将 xk作为稳态数据, 建立该稳态过程对应的新的稳态值 X; 如此继续。
对于稳态值输出, 可以根据判断结果, 仅输出在稳态过程结束时对应 的稳态值 X, 也可以根据远程需要, 将稳态过程中不同时间对应的稳态值 X输出。
在一个实施例中, 本发明还包括可以记录步骤, 从而对稳态过程的开 始时间 (对应暂态过程的结束时间) 、 暂态过程的开始时间 (对应稳态过 程的结束时间)进行记录, 并且记录在新的暂态过程开始前其相邻上一个 稳态过程的稳态值, 以及该暂态过程内的釆样值。
以上输出和记录的稳态数据比现有的间隔 T时间输出一次数据获得的 数据量少 (多个 T都在一个稳态过程)且避免了混迭错误。 并且, 它们真 实反映了系统在该时间段上的特性, 从而可作为后续系统计算的无误输 入。
由于稳态时间较暂态时间要长得多, 所以输出或传送的数据大幅度减 少。 另外, 又保证了信号不失真地被记录下来。 当然, 由记录的数据也能 够复现原信号。
在一个实施例中, 还可以包括预处理步骤, 用于对釆样出的釆样值进 行预处理, 然后直接或重抽样输出用于后续判断和计算。 一般地, 预处理 包括标度变换, 还可以包括坏数据剔除, 而对于正弦量, 例如电压、 电流、 功率等, 还可以包括有效值的计算。 在这种情况下, 输出的釆样值是有效 值。
需要指出的是, 对于轮询系统, 例如在规定的时间间隔 t (例如电力调 度自动化系统中 t = 3秒)或规定的时刻 (例如用电监视系统中每个整点开 始的每刻钟) , 需要釆样终端输出一次数据的情况下, 本发明的测量装置 和方法可以不直接输出稳态过程开始和结束的时标, 即, 时标不明确输出 而是暗含在轮询中。 当输出的数据在稳态过程时, 仅输出对应所需的稳态 值 X即可。 这种变化仍落在本发明的范围内。
本发明可以应用于热工、 电力系统中的温度、 压力、 流量、 电压、 电 流、 功率和相角等测量和记录中, 能够真实、 清晰地反映系统参数的变化 情况, 并且显著减少记录和输出数据的数量。
尽管已经示出和描述了本发明的实施例, 对于本领域的普通技术人员 而言, 可以理解在不脱离本发明的原理和精神的情况下可以对这些实施例 进行多种变化、 修改、 替换和变型, 本发明的范围由所附权利要求及其等 同限定。

Claims

权利要求书
1. 一种连续物理量测量装置, 其特征在于, 所述测量装置包括: 定时单元, 所述定时单元输出时基和时标;
釆样单元, 在所述时基控制下以满足奈奎斯特釆样定理的时间间隔对 连续物理量进行釆样, 并输出釆样值 xk;
判断单元, 判断釆样值 xk是否从稳态过程进入暂态过程或者是否从暂 态过程进入稳态过程;
计算单元, 在稳态过程中对所述釆样单元的釆样值计算稳态值 X; 输出单元, 根据判断结果以及所述时标, 输出稳态过程的开始时间和 所述稳态值 X , 以及暂态过程的开始时间和所述暂态过程的釆样值 xk
2. 如权利要求 1所述的测量装置, 其特征在于, 所述判断单元根据统 计法的 t分布或者滤波器输出进行判断。
3. 如权利要求 1所述的测量装置, 其特征在于, 还包括预处理单元, 和 /或有效值计算。
4. 如权利要求 1所述的测量装置, 其特征在于, 所述计算单元根据均 值或低通滤器计算所述稳态值。
5. 如权利要求 1所述的测量装置, 其特征在于, 还包括记录单元, 所 述记录单元对所述稳态过程的开始时间、 所述暂态过程的开始时间、 所述 暂态过程的开始时间前对应的稳态值以及所述暂态过程的釆样值进行记 录。
6. 一种连续物理量测量方法, 其特征在于, 所述测量方法包括以下步 骤:
在定时器输出时基的控制下以满足奈奎斯特釆样定理的时间间隔对连 续物理量进行釆样, 并输出釆样值 xk和时标;
判断釆样值 ¾是否从稳态过程进入暂态过程或者是否从暂态过程进入 稳态过程; 以及
根据判断结果以及所述时标, 输出稳态过程的开始时间和在稳态过程 中对釆样值计算获得的稳态值 X, 以及暂态过程的开始时间和所述暂态过 程的釆样值 ¾。
7. 如权利要求 6所述的测量方法, 其特征在于, 所述稳态值 X根据均 值或低通滤波器计算获得。
8. 如权利要求 6所述的测量方法, 其特征在于, 所述判断步骤包括: 计算釆样值 ¾对应的均值 和方差^ ;
判断(xt - ) / 是否服从 t分布;
若服从, 判断釆样值 xk未从稳态过程进入暂态过程或者判断釆样值 xk 从暂态过程进入稳态过程;
若不服从, 判断釆样值 xk从稳态过程进入暂态过程或者判断釆样值 xk 未从暂稳态过程进入稳态过程。
9. 如权利要求 8所述的测量方法, 其特征在于, 所述 t分布判断依据 以下公式:
Figure imgf000018_0001
其中"为风险系数。
10. 如权利要求 6所述的测量方法, 其特征在于, 所述判断步骤包括: 计算釆样值 ¾对应的均值 和方差^ ;
判断是否满足 |xt - |≤ ^ , 其中 A为给定值;
若满足, 判断釆样值 xk未从稳态过程进入暂态过程或者判断釆样值 xk 从暂态过程进入稳态过程;
若不满足, 判断釆样值 xk从稳态过程进入暂态过程或者判断釆样值 xk 未从暂稳态过程进入稳态过程。
11. 如权利要求 10所述的测量方法, 其特征在于, 所述 A位于 3 ~ 10 之间。
12. 如权利要求 6所述的测量方法, 其特征在于, 所述判断步骤包括: 计算釆样值 ¾对应的均值 ;
判断是否满足 |xt _ ft|≤ x„, 其中 δ为给定值, χη为所述连续物理量对 应的额定值;
若满足, 判断釆样值 xk未从稳态过程进入暂态过程或者判断釆样值 xk 从暂态过程进入稳态过程; 若不满足, 判断釆样值 xk从稳态过程进入暂态过程或者判断釆样值 xk 未从暂稳态过程进入稳态过程。
13. 如权利要求 12所述的测量方法, 其特征在于, 所述 δ位于 2%到 10%之间。
14. 如权利要求 6所述的测量方法, 其特征在于, 所述判断步骤包括: 对釆样值 ¾进行滤波, 分别得到釆样值 xk的滤波分量;
分别判断所述滤波分量是否超出对应设定的上限值; 以及
若均未超出, 判断釆样值 ¾未从稳态过程进入暂态过程;
若存在一个分量超出, 判断釆样值 xk从稳态过程进入暂态过程。
15. 如权利要求 6所述的测量方法, 其特征在于, 所述判断步骤包括: 对釆样值 ¾进行滤波, 分别得到釆样值 xk的滤波分量;
分别判断所述滤波分量是否超出对应设定的下限值; 以及
若存在一个分量超出, 判断釆样值 xk未从暂态过程进入稳态过程; 若均未超出, 判断釆样值 ¾从暂态过程进入稳态过程。
16. 如权利要求 6所述的测量方法, 其特征在于, 还包括对釆样值 xk 进行标度变换、 坏数据剔除、 重抽样和 /或有效值计算的预处理步骤。
17. 如权利要求 6所述的测量方法, 其特征在于, 还包括对所述稳态 过程的开始时间、 所述暂态过程的开始时间、 在所述暂态过程开始时间前 对应的稳态值以及所述暂态过程的釆样值进行记录的步骤。
PCT/CN2010/075008 2009-07-08 2010-07-06 连续物理量测量装置及方法 WO2011003347A1 (zh)

Priority Applications (3)

Application Number Priority Date Filing Date Title
US13/382,918 US8954294B2 (en) 2009-07-08 2010-07-06 Measuring device and measuring method for continuous physical quantity
CA2767586A CA2767586C (en) 2009-07-08 2010-07-06 Measuring device and measuring method for continuous physical quantity
EP10796724.2A EP2453213B1 (en) 2009-07-08 2010-07-06 Measuring device and measuring method for a continuous physical quantity

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN200910158375.X 2009-07-08
CN 200910158375 CN101614555B (zh) 2009-07-08 2009-07-08 连续物理量测量装置及方法

Publications (1)

Publication Number Publication Date
WO2011003347A1 true WO2011003347A1 (zh) 2011-01-13

Family

ID=41494332

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2010/075008 WO2011003347A1 (zh) 2009-07-08 2010-07-06 连续物理量测量装置及方法

Country Status (5)

Country Link
US (1) US8954294B2 (zh)
EP (1) EP2453213B1 (zh)
CN (1) CN101614555B (zh)
CA (1) CA2767586C (zh)
WO (1) WO2011003347A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103149485A (zh) * 2013-03-13 2013-06-12 绍兴电力局 一种基于电子式互感器的暂态波形分析系统

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101614555B (zh) * 2009-07-08 2011-05-18 保定市三川电气有限责任公司 连续物理量测量装置及方法
GB2488092B (en) * 2010-11-03 2014-10-29 Kittiwake Developments Ltd A sensor based means of monitoring the mechanical condition of rotating machinery that operates intermittently
CN102313834B (zh) * 2011-06-09 2013-10-09 保定市三川电气有限责任公司 一种电力数据的转换系统、方法与装置
CN102393214B (zh) * 2011-06-09 2014-11-05 郝玉山 连续物理量数据采集方法和装置
JP6215527B2 (ja) * 2012-02-02 2017-10-18 旭化成エレクトロニクス株式会社 物理量計測装置及び物理量計測方法
CN102928014B (zh) * 2012-10-23 2015-05-13 保定市三川电气有限责任公司 电力系统数字测量或遥测处理的方法及装置
CN103605023B (zh) * 2013-11-20 2016-09-21 国家电网公司 一种合并单元时间特性测量方法及测量装置
CN107797031B (zh) * 2016-08-30 2021-01-19 上海复旦微电子集团股份有限公司 电弧故障的检测方法及装置
CN106208248B (zh) * 2016-08-31 2019-07-23 浙江长兴笛卡尔科技有限公司 一种电动汽车bms的正时采集方法
CN109029522B (zh) * 2018-06-20 2020-10-27 泰州逸群信息科技有限公司 一种用于生物统计测量的情境化的记录设备

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH07244523A (ja) * 1994-03-02 1995-09-19 Toshiba Corp 定常状態検出装置
US6944554B2 (en) * 2003-07-11 2005-09-13 Sungkyunkwan University Method for detecting fault on transmission lines by using harmonics and state transition diagram
CN101216519A (zh) * 2007-12-29 2008-07-09 江苏省电力公司常州供电公司 配网电缆线路带电作业的测量方法
CN101459334A (zh) * 2007-12-14 2009-06-17 山东科汇电力自动化有限公司 电力系统故障信息获取方法
CN101614555A (zh) * 2009-07-08 2009-12-30 保定市三川电气有限责任公司 连续物理量测量装置及方法
CN101614554A (zh) * 2009-07-08 2009-12-30 保定市三川电气有限责任公司 连续物理量测量装置及方法

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5870436A (en) * 1997-01-02 1999-02-09 Raytheon Company Uniform discrete fourier transform filter parameter encoder
EP1965481A2 (en) * 2007-02-28 2008-09-03 STMicroelectronics, Inc. Integrated circuit and method for monitoring and controlling power and for detecting open load state

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH07244523A (ja) * 1994-03-02 1995-09-19 Toshiba Corp 定常状態検出装置
US6944554B2 (en) * 2003-07-11 2005-09-13 Sungkyunkwan University Method for detecting fault on transmission lines by using harmonics and state transition diagram
CN101459334A (zh) * 2007-12-14 2009-06-17 山东科汇电力自动化有限公司 电力系统故障信息获取方法
CN101216519A (zh) * 2007-12-29 2008-07-09 江苏省电力公司常州供电公司 配网电缆线路带电作业的测量方法
CN101614555A (zh) * 2009-07-08 2009-12-30 保定市三川电气有限责任公司 连续物理量测量装置及方法
CN101614554A (zh) * 2009-07-08 2009-12-30 保定市三川电气有限责任公司 连续物理量测量装置及方法

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
"Probability Theory and Mathematical Statistics", CHINA STATISTICS PUBLISHING HOUSE
"Probability Theory and Mathematical Statistics", ZHEJIANG UNIVERSITY

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103149485A (zh) * 2013-03-13 2013-06-12 绍兴电力局 一种基于电子式互感器的暂态波形分析系统

Also Published As

Publication number Publication date
CA2767586C (en) 2016-06-07
EP2453213B1 (en) 2018-06-20
EP2453213A1 (en) 2012-05-16
CA2767586A1 (en) 2011-01-13
US8954294B2 (en) 2015-02-10
CN101614555B (zh) 2011-05-18
EP2453213A4 (en) 2017-07-19
CN101614555A (zh) 2009-12-30
US20120109586A1 (en) 2012-05-03

Similar Documents

Publication Publication Date Title
WO2011003347A1 (zh) 连续物理量测量装置及方法
US8082367B2 (en) Differential time synchronization of intelligent electronic devices
EP3919899A1 (en) Formaldehyde concentration measurement method and apparatus, and air purifier
EP2690451A1 (en) System frequency measurement method, synchrophasor measurement method and device thereof
US20170030958A1 (en) Transformer parameter estimation using terminal measurements
JPWO2008126240A1 (ja) 同期フェーザ測定装置及びこれを用いた母線間位相角差測定装置
US8489351B2 (en) System and method for frequency measurement based on quadratic forms
CN109946513B (zh) 基于网络采样绝对延时补偿的跨间隔数字化计量方法、装置及系统
CN109444537A (zh) 一种计及带外干扰的自适应同步相量测量方法
WO2012167749A1 (zh) 连续物理量数据采集方法和装置
CN116298741B (zh) 一种绝缘体局部放电检测方法
CN107703342A (zh) 一种高精度暂态分析仪及其暂态算法
CN109298361B (zh) 基于暂态录波型故障指示器的故障分析预处理方法
CN115575880B (zh) 一种电力互感器计量误差状态在线检测方法
CN105242225B (zh) 一种动态相量测量装置的校准系统及其校准方法
CN115144052A (zh) 一种水电站水头计算方法及系统
CN108957174A (zh) 一种电压暂降检测装置及方法
CN103630744A (zh) Pmu相角检测方法及系统
CN112180314A (zh) 一种抗干扰自校正gis电子式互感器采集方法和装置
CN110780114A (zh) 一种台区理论线损分析方法及可读存储介质
CN110187198B (zh) 一种频率器件性能评估的方法及装置
CN102832911A (zh) 一种数字信号恢复方法及装置
TW201224473A (en) Electricity feature identification device and method thereof
CN104280601B (zh) 一种车辆智能监测用光伏电源的电压测量系统
CN109444785B (zh) 一种数字量报文准确度的测量方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 10796724

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 13382918

Country of ref document: US

NENP Non-entry into the national phase

Ref country code: DE

WWE Wipo information: entry into national phase

Ref document number: 2767586

Country of ref document: CA

WWE Wipo information: entry into national phase

Ref document number: 2010796724

Country of ref document: EP