CN105117504A - Improved revolving door data acquisition method used for signal system - Google Patents

Abstract

The present invention relates to an improved revolving door data acquisition method used for a signal system. The method comprises the following steps of: (1) setting a sampling frequency to perform data acquisition; (2) calculating a slope threshold value of a current datum and a previous stored datum; (3) calculating a slope of a next acquired datum and the previous stored datum, and comparing the slope with the slope threshold value; and (4) if the slope is in the range of the slope threshold value, rejecting the current datum, updating the rejected current datum as the current datum and returning to the step (2), or otherwise, storing the current datum, updating the stored current datum into the previous stored datum, updating the next acquired datum into the current datum and returning to the step (2). Compared to the prior art, with the adoption of the method provided by the present invention, a data acquisition speed is increased, storage devices are reduced, the acquisition precision of action change data is improved, the acquisition quality is improved, the acquisition precision can be controlled according to a required error range, and the algorithm performance is improved as far as possible.

Description

A kind of modified revolving door collecting method for signal system

Technical field

The present invention relates to a kind of collecting method of signal system, especially relate to a kind of modified revolving door collecting method for signal system.

Background technology

In Rail Transit System, signal system is a very important Mechatronic Systems integrating train traffic control and Train Detection and Identification, its core is train automatic controlling system, realize ground to control and car controls to combine, local control combines with center-control, form one based on safety equipment, collection train traffic control, the automatic control system that the functions such as combustion adjustment and train driving robotization are integrated.A large amount of memory device must be utilized at present to carry out the on-site supervision data of saved system stable operation, if complete preservation, because information redundance is high, take too much memory device resource, make the collection of data and transfer efficiency low, the method of data compression can be adopted in order to accelerate acquisition speed, mainly can be divided into two large classes, lossy compression method and Lossless Compression, Lossless Compression utilizes the statistical redundancy of data to compress, data can be recovered completely and not cause any distortion, but compressibility is not high, in supervisory system, in order to accurate historical data values can be obtained, the overwhelming majority needs lossy compression method, lossy compression method allows in data acquisition, lose certain information, but the information of loss is little to the informational influence understanding historical data, such energy obtains higher ratio of compression, save a large amount of storage spaces.

Revolving door data compression algorithm is a kind of compression algorithm of linear trend, and emphasis is it is considered that the slope variation situation of linear trend, and this algorithm logic is simple, and execution efficiency is high, obtains and quote widely in supervisory system real-time data acquisition.The main thought of revolving door algorithm utilizes current data value and the previous data value stored to form the limited field of compression to judge whether current data should be given up, if Current data values is in compression limited range, then data point is given up, if outside scope, then preserves.Although this algorithm has higher ratio of compression, choosing of parameter is fixing or dependence experience and experiment, directly can affect the performance of algorithm like this.

Summary of the invention

Object of the present invention be exactly in order to overcome above-mentioned prior art exist defect and a kind of modified revolving door collecting method for signal system is provided, under the prerequisite ensureing effective information data, maximum raising information acquisition ratio of compression, thus improve the efficiency of data acquisition.

Object of the present invention can be achieved through the following technical solutions:

For a modified revolving door collecting method for signal system, the method is used for fast-changing switching value and analog quantity in collection signal system, and it is characterized in that, described collecting method comprises the following steps:

(1) setting sensor sample frequency carry out data acquisition;

(2) first image data directly preserved as upper preservation data, second image data is as current data;

(3) current data and a upper slope threshold values preserved between data is calculated;

(4) read in next image data, calculate next image data and a upper slope preserving data, if in slope threshold range, then perform step (5), otherwise perform step (6);

(5) give up current data, judge whether next image data is last collection point, if then directly preserve and end data collection, otherwise be updated to new current data, and return step (3);

(6) current data is preserved, judge whether next image data is last collection point, if then directly preserve and end data collection, otherwise it is updated to upper one and preserves data by the current data of preservation, next image data is updated to current data, and returns step (3).

Described step (3) specifically comprises following sub-step:

(301) one is set to preserve data as D (x _saved, y _saved), current data is D (x _now, y _now), threshold threshold is d;

(302) D (x is calculated _now, y _now) and D (x _saved, y _saved) slope k up and down that forms _upand k _down:

$k_{up}=\frac{y_{now}-(y_{saved}-d)}{x_{now}-x_{saved}}$

$k_{down}=\frac{y_{now}-(y_{saved}+d)}{x_{now}-x_{saved}}.$

In described step (4), the next image data of calculating and a upper slope preserving data are specially: set next image data as D (x _next, y _next), calculate next image data and a upper slope preserving data is:

$k_m=\frac{y_{next}-(y_{saved}-d)}{x_{next}-x_{saved}}.$

Described threshold threshold d regulates according to the data feedback after collection.

Described threshold threshold d regulates according to the data feedback after collection, specifically comprises the following steps:

(401) obtaining original data series is { D ₁, D ₂, D ₃... D _n, and to reduce the new ordered series of numbers obtained after carrying out again decompress(ion) to the data after gathering be { D ₁', D ₂', D ₃' ... D _n';

(402) standard variance δ is calculated:

$\mathrm{\δ}=\sqrt{\frac{1}{n}{\mathrm{\Σ}}_{i=1}^n{(D_i-{D_i}^{,})}^2};$

(403) according to the error limited range μ of data acquisition _minand μ _maxthreshold threshold d is adjusted:

When time, then the threshold threshold d starting to set is constant, gathers effectively;

When time, new threshold threshold d '=d+ δ, this time collection result is invalid, by new threshold threshold collection to meeting error range;

When time, new threshold threshold d '=d-δ, this time collection result is invalid, by new threshold threshold collection to meeting error range.

Compared with prior art, tool of the present invention has the following advantages:

(1) during device standby in signal system, analog quantity change steadily, only have the change when there being action violent, the most of the time is standby, adopts modified revolving door collecting method to compress redundant data as much as possible, improve acquisition speed, decrease memory device.

(2) in signal system, analog quantity is when action, and data variation is violent, adopts modified revolving door collecting method can improve the acquisition precision of action delta data, improves acquisition quality.

(3) by the data feedback after collection, the threshold threshold parameter of adjustment revolving door, the error range that can realize as requested controls acquisition precision, improves algorithm performance to greatest extent.

Accompanying drawing explanation

Fig. 1 is a kind of modified revolving door collecting method process flow diagram for signal system of the present invention.

Embodiment

Below in conjunction with the drawings and specific embodiments, the present invention is described in detail.

Embodiment

A kind of modified revolving door collecting method for signal system of the present invention, this collecting method be used in tramcar light current integrated in, the comprehensive monitoring system of its to be a kind of with train traffic control be core, achieves the integrated of the weak electricity system such as ATO supervisory system ATS, electric power monitoring system PSCADA, Customer information display system PIS, rolling stock section broadcast system PA, closed-circuit television system CCTV, the limited control in crossing.Train Dispatching, scheduling such as multisystem such as electricity tune, ring tune, highway communication etc. are organically combined, thus is better driving service.This system sets up a unified hardware and software platform and man-machine interface, integrated by each system, defining with train traffic control is the integration scheduling of core, towards car business and the information control platform of safeguarding unity, towards passenger's public information interactive service management function.Wherein in ATS, the main information gathered is that orbit occupancy state, route state, train operation state and the signalling arrangement fault etc. that trackside and vehicle-mounted ATP provide control and supervise the Back ground Information (track section, track switch, teleseme operation information) of train operation; In PSCADA, the main information gathered is the data such as rectification alternating temperature, power alternating temperature, rail potential; In PIS system, the main information gathered is the data such as real-time train operation information, audio/video information; In CCTV system, the main information gathered is audio and video frequency signal.

Fig. 1 is collecting method process flow diagram of the present invention, is described in detail below to each step in figure:

In step 1, need the sample frequency of setting sensor according to system;

In step 2, gather analog quantity and switching value information according to special or standard industry communication agreement;

In step 3, use revolving door algorithm, compression process is carried out to each switching value and analog quantity information, first preserves start point data, be set to D (x _saved, y _saved), second data is current data, is set to D (x _now, y _now);

In step 4, if the threshold threshold of revolving door is d, current data D (x _now, y _now) and a upper preservation data D (x _saved, y _saved) slope value up and down that forms is k _upand k _down, the formula calculating slope is as follows:

$k_{up}=\frac{y_{now}-(y_{saved}-d)}{x_{now}-x_{saved}}$

$k_{down}=\frac{y_{now}-(y_{saved}+d)}{x_{now}-x_{saved}}$

In step 5, reading in next image data is D (x _next, y _next), then these data and a upper slope preserving data are k _m, k _mcomputing formula as follows:

$k_m=\frac{y_{next}-(y_{saved}-d)}{x_{next}-x_{saved}}$

By three slope value k in step 6 _up, k _downand k _mcompare, if meet k _up>=k _m>=k _down, then perform step 7, otherwise perform step 11;

In step 7, current data is given up and performs step 8;

Next data point D (x is judged in step 8 _next, y _next) whether be last data point, if then perform step 9, directly preserve and end data collection in step 9; If result of determination is no in step 8, then perform step 10, by next image data D (x in step 10 _next, y _next) assignment is new D ' (x _now, y _now) and return step 4, now slope should restart to calculate, and new slope is k _up', k _down':

Continue to perform step 5;

In step 11, by current data point D (x _now, y _now) preserve, perform step 12;

In step 12, judge next data point D (x _next, y _next) whether be last data point, if then perform step 9, directly preserve and end data collection; If result of determination is no in step 12, then perform step 13;

Be a new upper preservation data point D ' (x by the current data point assignment of having preserved in step 13 _saved, y _saved), by next image data D (x _next, y _next) assignment is new current data D ' (x _now, y _now), ensuing executive routine returns step 4, performs successively.

For choosing of threshold threshold d, by the training of historical data, each analog quantity is trained.Its process is as follows:

After carrying out the revolving door data acquisition be in summary of the invention, if original ordered series of numbers is { D ₁, D ₂, D ₃... D _n, and to reduce the new ordered series of numbers obtained after carrying out again decompress(ion) to the data after gathering be { D ₁', D ₂', D ₃' ... D _n', calculating standard variance is δ, and its computing formula is:

$\mathrm{\δ}=\sqrt{\frac{1}{n}{\mathrm{\Σ}}_{i=1}^n{(D_i-{D_i}^{,})}^2};$

Generally can there is limited range μ to the error of data acquisition _minand μ _max, perform different statements according to the condition of three below:

When time, then the threshold threshold d starting to set is constant; Then gather effectively;

When time, new threshold threshold d '=d+ δ; This time collection result is invalid, by new threshold threshold collection to meeting error range;

When time, new threshold threshold d '=d-δ; This time collection result is invalid, by new threshold threshold collection to meeting error range;

After said process, d value, after feedback adjusting, can meet within the scope of data acquisition errors, maximum raising ratio of compression, improves algorithm performance.

Before using algorithm, gathering density is five points per second, and after algorithm, ratio of compression is 26.71, surplus storage space 96.26%, the more important thing is five points per second before gathering density crucial moment (during as track switch action) has, becomes tens points.

After using this algorithm, sensor can carry out collect and transmit with maximum speed.

The field apparatus of access, before using algorithm, self-regulating machine accesses tens field apparatuss, due to upstream bandwidth restriction (being 20MHz to the maximum), can access now up to a hundred.

As can be seen here, this Quick Acquisition method for mobile unit collection vehicle information and platform equipment collection signal system (as information such as track switch action current, teleseme electric currents) substantially increases collecting efficiency.

Claims (5)

1., for a modified revolving door collecting method for signal system, the method is used for fast-changing switching value and analog quantity in collection signal system, and it is characterized in that, described collecting method comprises the following steps:

(1) setting sensor sample frequency carry out data acquisition;

(2) first image data directly preserved as upper preservation data, second image data is as current data;

(3) current data and a upper slope threshold values preserved between data is calculated;

(4) read in next image data, calculate next image data and a upper slope preserving data, if in slope threshold range, then perform step (5), otherwise perform step (6);

(5) give up current data, judge whether next image data is last collection point, if then directly preserve and end data collection, otherwise be updated to new current data, and return step (3);

(6) current data is preserved, judge whether next image data is last collection point, if then directly preserve and end data collection, otherwise it is updated to upper one and preserves data by the current data of preservation, next image data is updated to current data, and returns step (3).

2. a kind of modified revolving door collecting method for signal system according to claim 1, it is characterized in that, described step (3) specifically comprises following sub-step:

(301) one is set to preserve data as D (x _saved, y _saved), current data is D (x _now, y _now), threshold threshold is d;

(302) D (x is calculated _now, y _now) and D (x _saved, y _saved) slope k up and down that forms _upand k _down:

3. a kind of modified revolving door collecting method for signal system according to claim 2, it is characterized in that, in described step (4), the next image data of calculating and a upper slope preserving data are specially: set next image data as D (x _next, y _next), calculate next image data and a upper slope preserving data is:

4. a kind of modified revolving door collecting method for signal system according to claim 3, is characterized in that, described threshold threshold d regulates according to the data feedback after collection.

5. a kind of modified revolving door collecting method for signal system according to claim 4, is characterized in that, described threshold threshold d regulates according to the data feedback after collection, specifically comprises the following steps:

(401) obtaining original data series is { D ₁, D ₂, D ₃d _n, and to reduce the new ordered series of numbers obtained after carrying out again decompress(ion) to the data after gathering be { D ₁', D ₃', D ₃' ... D _n';

(402) standard variance δ is calculated:

(403) according to the error limited range μ of data acquisition _minand μ _maxthreshold threshold d is adjusted:

When time, then the threshold threshold d starting to set is constant, gathers effectively;

When time, new threshold threshold d '=d+ δ, this time collection result is invalid, by new threshold threshold collection to meeting error range;

When time, new threshold threshold d '=d-δ, this time collection result is invalid, by new threshold threshold collection to meeting error range.