CN101520787B - Method for storing real-time data - Google Patents

Abstract

A method for storing real-time data, the collector collects the data of the storage sensor in real time; reads the summary information of the main part index and verifies and compares it; extracts the latest pointer of the main part index content; reads the latest storage pointer of the data content; stores the data Content and check code; calculate new index content and check code; store primary index content and check code; update and store primary index summary; store backup index content and check code; store backup index summary; read backup Index summary, check and compare the backup index summary; obtain the latest pointer of the backup index; read the latest storage pointer of the data content; re-establish the storage information. Double index and multiple checks are used to protect the data in the memory to ensure the integrity of the data stored in the collector, and cyclic coverage is used to realize continuous storage of data content under fixed storage capacity, improving the reliability and integrity of important parameter records in industrial processes Sex, improve the work efficiency of process monitoring and safety production.

Description

A method for storing real-time data

technical field

The present invention relates to the field of information processing technology in the field of information technology and automation, and relates to a method for storing real-time data used on an embedded collector, which is a software process executed on an embedded operating system and is suitable for harsh application environments Embedded collector real-time data storage.

Background technique

In the industrial production process, the change process of various parameters often reflects the operation status of the production equipment and the connection of the production process in a stage. Requirements, etc. are of great significance. Embedded collectors are widely used in the industrial field to measure and record many industrial parameters (such as temperature, pressure, flow, etc.), generate continuous historical data, and provide first-hand information for analysis and research for technicians and managers. However, due to the harsh working environment in many industrial productions (such as high temperature, high humidity, strong corrosion, etc.), the operating conditions of the embedded collector are not very stable (such as frequent power failures, severe external However, it is still unavoidable that accidents such as power failure, restart, or even crash and damage of the collector will occur frequently. In this way, if the program is accidentally interrupted during the storage process, the data stored in the memory will be incomplete, and the data content may not be able to be interpreted when it is read again. In order to ensure that the stored data in the collector memory is not lost under the condition that the hardware of the memory is not damaged, it is necessary to adopt a high-reliability data storage method to improve the utilization rate of the memory, improve the fault tolerance of the software system, and ensure the continuity of historical data. sex. At present, the data storage on the computer uses a standard file management system, such as FAT, NTFS, etc., which all adopt a directory-based management method. In the actual physical address, the storage address of the file content is pointed to the storage address of the file content step by step through the pointer. Unified to many The data of each file is managed, so that once a certain directory pointer in the physical address is wrong, the content in the file cannot be read and interpreted, and it is risky to use it in an embedded collector. Many current embedded collectors use a record-based storage method. Each record contains all the information of the data content and is stored continuously in the physical address to ensure that all data is unreadable due to some data corruption. However, this method requires each record to adopt a fixed storage format to contain all storage information, causing each record to take up many bytes, which is not convenient for storing more records in a memory with limited capacity. Therefore, a high-efficiency, high-reliability data storage method is an effective means to solve the data storage requirements of embedded collectors in the industrial production process.

Contents of the invention

In order to solve the problems of the prior art, the purpose of the present invention is to provide a method for storing real-time data in an embedded collector, which can efficiently and reliably store the parameter change process in industrial process monitoring, and avoid failure due to external factors. All the stored historical data cannot be read, improving the safety of embedded collectors and ensuring the parameter monitoring process in industrial production.

In order to achieve the above object, the technical solution of the present invention is to provide a method for storing real-time data by an embedded collector based on a double index, and its technical solution is:

Step A: first collect the values of each sensor in real time by the collector, and obtain the data content to be stored after processing;

Step B: Read the summary information of the primary share index;

Step C: Verify and compare the summary information of the primary share index; if the summary information of the primary share index is verified to be correct, perform step D; if the summary information of the primary share index is verified to be wrong, perform step N;

Step D: Extract the latest pointer of the main index content;

Step E: Read the latest storage pointer of the primary index content, and use the latest storage pointer to read the latest piece of index information in the index content;

Step F: Verify and compare the latest index information, if the verification is wrong, execute step N, if the verification is correct, execute step G;

Step G: extract the latest storage pointer of the data content of step Q from the latest index information, and store the data content and check code of the hour, minute, second, channel, and value in the address where the storage pointer is located;

Step H: Calculate the new index content and check code, and decide whether to perform circular overwriting operation according to whether the storage pointer reaches the physical maximum storage address. If overwriting is required, the index year, month, day, storage content start address, Content-length content is adjusted accordingly;

Step 1: store the main part index content and check code;

Step J: update the summary of the primary and share index;

Step K: storing the master-part index summary;

Step L: storing backup index content and check code;

Step M: Store the backup index summary, complete the storage operation, and start re-execution from step A;

Step N: Read the summary of the backup index, and use the backup index to perform related operations;

Step O: When operating the backup index summary, perform a verification comparison. If the content verification error in the backup index summary indicates that the data in the memory has been completely damaged, and the software cannot continue to maintain it, then perform step R; if the backup index The content in is also verified to be correct, then execute step P;

Step P: Obtain the latest pointer of the backup index;

Step Q: read the latest storage pointer of the data content, and execute step G;

Step R: re-establish storage information, and start from step A again.

According to an embodiment of the present invention, the real-time data collection also includes:

Step A1: The real-time collection includes: channel number, year, month, day, hour, minute, second, data value information, check code data to generate an index and data content on a daily basis;

Step A2: Compare the index information including the year, month, day, start address, end address, check code, and the stored index with the stored index to determine the storage address of the data content;

Step A3: Determine whether the stored data needs to be overwritten according to the storage address of the data content. If overwriting is required, delete the content in the overwritten storage area, store the data content in this area, and change the index information; if not, directly Store the data content in the storage address obtained by step A2;

Step A4: Determine whether the index information needs to be updated according to the storage situation of the data content, if necessary, change the stored index to new index information, and if not, keep the original index unchanged.

According to an embodiment of the present invention, the stored data content coverage adopts a cyclic coverage storage method of circular sequential storage, and the steps are as follows:

Step A31: sequentially store in the SD memory card according to the physical address sequence;

Step A32: When storing to the maximum physical address of the SD memory card, delete the data of the earliest day in the SD memory card;

Step A33: Then continue to store data at the physical address where the stored content was stored on the earliest day to form a circular storage mode, and make corresponding changes to the index content.

According to an embodiment of the present invention, the index also includes:

The index includes the physical storage address pointing to the data content of the day, and the storage index adopts the storage mode of the main part and the backup of the double index, and the main part index and the backup index with the same content are stored in different physical addresses in the ferroelectric memory, Each contains verification information;

The storage index consists of an index summary and index content. The content of the index summary includes the starting and ending physical addresses of the index content, the total number of storage records, and the remaining space of the physical storage. The index content is the index information pointing to the stored data generated in units of days.

According to an embodiment of the present invention, in the real-time data storage method of the embedded collector, the check code R (X) polynomial is a CRC16 cyclic check code, and its calculation formula is

$\frac{\mathrm{<span\; class="notranslate">\; B</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; 2</span>}}^{\mathrm{<span\; class="notranslate">\; 16</span>}}}{\mathrm{<span\; class="notranslate">\; G</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; Q</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; G</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span>}$

Wherein B(X) is the n-bit binary sequence number to be checked, G(X)=X ¹⁶ +X ¹⁵ +X ² +1 is a fixed polynomial, Q(X) is an integer polynomial, and the remainder of R(X) is Cyclic check code, X is a polynomial free variable.

According to an embodiment of the present invention, in the real-time data storage method of the embedded collector, the real-time data value is a single-precision floating-point number, represented by 4 bytes, and the start address and the end address are integer numbers, represented by 4 bytes representation.

According to an embodiment of the present invention, in the real-time data storage method of the embedded collector, the storage capacity of the SD memory card is 256MB-4GB, and the storage capacity of the ferroelectric memory is 16KB-128KB.

According to an embodiment of the present invention, in the real-time data storage method of the embedded collector, the embedded collector is a device for transplanting an embedded operating system composed of an ARM processor, and the file system is not transplanted in the device software .

The method for storing real-time data of the present invention is applied to an embedded ARM collector.

The embedded collector proposed by the present invention is the main advantage of the reliable real-time data storage method used in the acquisition system of hardware as follows: the index mode is used to manage the data content, which can reduce the occupied bytes of each data content, and improve the The number of data records stored under a certain storage capacity; each data record is verified by CRC16 to ensure the correctness of each stored data; the storage method of primary and backup double indexes is adopted to avoid the An index is damaged and the entire data content cannot be explained; the primary and backup indexes are composed of index summary and index content, and each has a CRC16 check code, which ensures the maintainability of reading and interpretation of the stored content; data The storage process follows the order of content first, then index, first master copy, and then backup, which ensures that the stored information will not be confused due to unexpected interruption during the storage process. The continuous storage of data content under fixed storage capacity is achieved by means of cyclic coverage, and the earliest historical data is covered with the latest collected data to ensure the continuity of stored data in the collector, which can improve the reliability and integrity of important parameter records in the industrial production process . This storage method is suitable for the application of embedded collectors, and is especially suitable for use in industrial collection occasions with multiple interference environments and harsh conditions.

Description of drawings

Fig. 1 is the flowchart of the method for storing real-time data using an embedded collector in an embodiment of the present invention;

Fig. 2 embodiment of the present invention adopts the structural block diagram of the real-time data storage system of embedded collector;

Fig. 3 is a schematic diagram of an embodiment of the ring coverage mode in the real-time data storage method of the embedded collector of the present invention.

Detailed ways

The present invention will be described in detail below in conjunction with the accompanying drawings. It should be noted that the described embodiments are only intended to facilitate the understanding of the present invention, rather than limiting it in any way.

A method for storing real-time data using an embedded collector of the present invention, the data of multiple sensors collected in real time by the embedded collector is used as a storage object; it is a high-reliability ARM collector used in an industrial collection environment A dedicated method for real-time data storage, using a special storage format and management method, consisting of storage indexes and data content.

Each piece of real-time data collected by the collector includes channel number, year, month, day, hour, minute, second, and real-time data value information. The real-time data value is a single-precision floating-point number represented by 4 bytes. When storing, the ferroelectric memory and SD memory card are used as storage media. The storage index is composed of a primary index and a backup index. They have the same content, and each contains verification information, which is stored in the ferroelectric memory and points to the data content in the memory card. Both the primary index and the backup index are composed of an index summary and index content. The content of the index summary includes the starting and ending physical addresses of the index content, the total number of stored records, and the remaining space of the SD memory card. The index content is pointing data in units of days Content information, including year, month, day, data content start address and end address, where the start address and end address are represented by 4-byte integers. Each data record in the data content includes channel number, hour, minute, second, real-time data value, and check code information.

After the data collector collects real-time data, it adds the time information of the collector into it, making it a data record containing channel number, year, month, day, hour, minute, second, and data value information, and then divides it by day into a data record. The unit generates an index and data content, where the index includes year, month, day, and the data content includes channel number, hour, minute, second, and data value information; compare the index with the latest stored index to determine the data content According to the storage address of the data content, it is judged whether the stored data needs to be overwritten. If overwriting is required, the content of the overwritten storage area is deleted, the data content is stored in this area, and the index information is changed. The specific overwriting method As described in the embodiment of ring coverage as shown in Figure 3; if not needed, then directly store the data content in the storage address obtained from the storage index; determine whether the index information needs to be updated according to the storage situation of the data content, and if necessary, The stored index is changed to the new index information, and the original index remains unchanged if it is not needed.

Fig. 1 is a flow chart of the real-time data storage method using the embedded collector, which describes the entire process of data storage of the embedded collector, and its specific steps are:

Step A: The value of each sensor is collected by the collector, and the data content to be stored is obtained after processing; Step B: Read the summary information of the master index; Step C: Verify and compare it using the CRC16 cycle check code , to determine whether to read the backup index information. The calculation formula of CRC16 cycle check code is:

$\frac{\mathrm{<span\; class="notranslate">\; B</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; 2</span>}}^{\mathrm{<span\; class="notranslate">\; 16</span>}}}{\mathrm{<span\; class="notranslate">\; G</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; Q</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; G</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span>}$

Wherein B(X) is the n-bit binary sequence number to be checked, G(X)=X ¹⁶ +X ¹⁵ +X ² +1 is a fixed polynomial, Q(X) is an integer polynomial, and the remainder of R(X) is cyclic checksum. The verification process is realized by circular shift and XOR operation, and its code is the same as the standard CRC16 operation code.

If the check code calculated by the primary share index summary information does not match the stored check code, it means that the check code is wrong, and then execute step N, and use the backup index to perform related operations. If they match, it means that the check code is correct, and then execute step D ;Step D: Extract the latest pointer of the main share index content from the master share index summary information; Step E: Use the latest storage pointer of the master share index to read the latest index information in the index content; Step F: Perform the latest index information Check and compare, if the check is wrong, execute step N, if the check is correct, execute step G; step G: extract the latest storage pointer of the data content of step Q from the latest index information, and store it in the address where the storage pointer is located Store data content, including hour, minute, second, channel number, real-time value and CRC16 check code; Step H: Calculate the new index content and check code, and decide whether to perform ring coverage operation according to whether the pointer reaches the physical maximum storage address , if it needs to be overwritten, adjust the year, month, day, storage content start address, and storage content length information in the index content accordingly, and use the overwriting method to replace the stored data with new data; if it does not need overwriting, directly replace the The data content is stored in the corresponding storage address; step I: store the main share index content and check code in the physical address pointed to by the latest main share index content pointer; step J: update the main share according to the information of the main share index content Index summary; step K: store the primary share index summary at the location of the original primary share index summary; step L: copy the same index content and verification code as the main share index content, and store it in the physical location pointed to by the latest backup index content In the address; Step M: Store the backup index summary, complete the storage operation, and re-execute from Step A; Step N: When using the backup index operation, perform CRC16 check comparison, if the content in the backup index is also verified incorrectly , it means that the data in the memory has been completely damaged, and the software method can no longer be used for maintenance, then perform step R; if the content in the backup index is verified to be correct, then perform step P; step P: get the backup from the backup index summary The latest pointer of the index content; step Q: read the latest storage pointer of the data content, and execute step G; step R: re-establish the storage information, and re-execute from step A.

With this storage method and storage order, it can be guaranteed that the data content can still be interpreted through the index when the software program is interrupted unexpectedly: if the content is interrupted when storing the content, since the index information has not been modified, there will be no error when interpreting the data ; If it is interrupted when storing the main part index content, since the main part index summary has not been modified, it can be interpreted according to the data before the content has not changed; if it is interrupted when storing the main part index summary, then check and compare At this time, an error will occur, and since the backup index information has not been modified at this time, the backup index information can be used to explain the data before the data content has not changed. In this way, as long as the physical storage medium is not damaged, according to this storage method, one of the primary index and the backup index must have complete and correct information, and the data content can be interpreted without causing a storage error due to a certain data record And cause the entire data content to be unreadable for interpretation.

The technical core of the present invention is to apply multiple verification and double index methods in the data storage process, so that the stored content has higher reliability, and can ensure that the embedded collector working in a harsh environment is not affected by external influences from the software. Reliable processing performance, enhance the maintainability of data storage, and improve the credibility of historical data. It uses reasonable execution steps to avoid unreadable data caused by unexpected interruptions, improves the utilization efficiency of storage media, and enhances the integrity of stored data. It is safer and more reliable than ordinary embedded file systems, ensuring physical The readability of the data content under the condition that the memory is not damaged is of great significance for the long-term monitoring and recording of important process parameters in the industrial production process.

In order to realize the method for storing real-time data of the present invention, the embodiment shows a structural block diagram of a typical real-time data storage system using an embedded ARM collector in Figure 2, and the real-time data storage system acquisition device is mainly composed of N sensors 1. An embedded ARM collector 2 and a computer 3 are composed, and N sensors are represented as 11, 12, 13...1N, wherein the embedded ARM collector 2 includes a computer communication interface 21, an ARM controller 22, an iron Electric memory 23, SD memory card 24, communication interface 25, the communication interface 25 can be RS232, RS485 or CAN interface, computer communication interface 21 can be RS232, USB or Ethernet interface, sensors 11, 12, 13... ... 1N is a digital sensor with the same communication interface 25, which is connected to the embedded ARM collector 2 through a communication bus. The storage capacity of the SD memory card is 256MB-4GB, and the storage capacity of the ferroelectric memory is 16KB-128KB. The embedded collector is a device for transplanting an embedded operating system composed of an ARM processor, and the common file system is not transplanted in the software of the device.

The working principle of Fig. 2 is as follows: the ARM controller 22 sequentially sends collection commands to each sensor 11, 12, 13...1N through the communication interface 25, and then receives the transmission from the sensors 11, 12, 13...1N numerical information, and generate complete real-time data content to be stored according to the clock information of the ARM controller 22 itself, and then decompose the real-time data content into index information and data content according to the method for storing real-time data of the invention, respectively It is stored in the ferroelectric memory 23 and the SD memory card 24 . When the computer 3 is connected to the embedded ARM collector 2 through the computer communication interface 21, the data content stored in the ferroelectric memory 23 and the SD memory card 24 can be read to obtain the historical records of the parameters monitored by the collector 2.

Fig. 3 is a schematic diagram of an embodiment of a real-time data storage method of an embedded collector in a circular coverage mode. In the figure of this embodiment, in order to simplify the description process, the primary index content and the backup index content are collectively referred to as index information, and the physical storage area for storing data content is managed in blocks, and each unit is called a data block. These names are used in this real-time example, and other names are the same as those in the claims, which are hereby described. The physical storage area is composed of m data blocks including BL1, BL2, ..., BLm. Each index information In contains the start address Sa and the end address Ea respectively. For the current latest index information, it contains the start address The address Sa and the latest content storage address N (referred to as the current address, the same below). When using the storage method of the present invention, three storage index information In1, In2, and In3 have been stored in the memory in the embodiment, and the data content is sequentially stored in the storage area according to the order of physical addresses from small to large, wherein the index In3 It is the latest index information. The start address Sa and end address Ea contained in each of them point to the corresponding position in the data area. The start address Sa1 points to the beginning of the data block BL1, and the end address Ea1 and start address Sa2 point to the data block. The same address of the block BL2, the end address Ea2 and the start address Sa3 point to the same address of the data block BL4, and the current address Na3 of the index In3 points to the bottom of the data block BLm (ie, the maximum physical storage address boundary). The corresponding relationship at this moment is marked with a solid line. After the collector collects a new piece of data, since the physical storage is full, the old data must be cleared before new data can be stored. At this time, the earliest data is the data block BL1 pointed to by the index In1. First, the program replaces the current address Na3 in the index In3 with the end address Ea3, then clears the content of the data block BL1, and then creates a new index In4, which starts The start address Sa4 and the current address Na4 point to the start of the data block BL1, and the start address Sa1 of the modified index In1 points to the start of the data block BL2. The corresponding relationship at this time is marked with a short dashed line. After collecting and storing for a period of time according to the data storage method of the present invention, the current address of the index In4 points to the bottom of the data block BL1 (marked by a long dotted line), and now the data is full again and needs to be overwritten. After collecting new data at this time, first clear the content in the data block BL2, and then modify the current address Na4 of the index In4 to point to the beginning of the data block BL2. At this time, because the end address Ea1 of the index In1 points to the data block BL2 In the middle, it means that there is no data of the index In1 in the memory, the index In1 is deleted from the memory, and the starting address Sa2 of the index In2 is modified to point to the starting end of the data block BL3. The corresponding relationship at this time is marked with a long dotted line. Using this cyclic coverage storage method can ensure that the memory is always used with high efficiency, and can ensure that the data stored in the memory is the latest data content, alleviating the impact of the memory size limit on the data storage capacity limit.

The above is only a specific implementation mode in the present invention, but the scope of protection of the present invention is not limited thereto. Anyone familiar with the technology can understand the conceivable transformation or replacement within the technical scope disclosed in the present invention. All should be covered within the scope of the present invention, therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims (10)

1. A method for storing real-time data is aimed at the design of an embedded ARM collector, storing a primary index and a backup index. The primary index and the backup index are respectively composed of an index summary and an index content, wherein the content of the index summary Contains the starting and ending physical addresses of the index content, and the total number of storage records; the index content is the information pointing to the data content in units of days, including year, month, day, data content start address and end address; it is characterized in that: Its storage steps are: Step A: first collect the data of each sensor in real time by the collector, and obtain the data content to be stored after processing; Step B: Read the summary information of the primary share index; Step C: Verify and compare the summary information of the primary share index; if the summary information of the primary share index is verified to be correct, perform step D; if the summary information of the primary share index is verified to be wrong, perform step N; Step D: Extract the latest storage pointer of the primary index content; Step E: Read the latest storage pointer of the primary index content, and use the latest storage pointer to read the latest piece of index information in the index content; Step F: Verify and compare the latest index information, if the verification is wrong, execute step N, if the verification is correct, execute step G; Step G: Extract the latest storage pointer of the data content from the latest index information, and store the data content and check code of the hour, minute, second, channel, and value in the address where the latest storage pointer of the data content is located; Step H: Calculate the new index content and check code, and decide whether to perform circular overwriting operation according to whether the latest storage pointer of the data content in step G reaches the physical maximum storage address. If overwriting is required, the index year, month, and day , the starting address of the storage content, and the content of the storage content length are adjusted accordingly; Step 1: store the main part index content and check code; Step J: update the master-part index summary; Step K: storing the master share index summary; Step L: storing backup index content and check code; Step M: Store the backup index summary, complete the storage operation, and start re-execution from step A; Step N: Read the summary of the backup index, and use the backup index to perform related operations; Step O: Verify and compare the backup index summary. If the content in the backup index summary is incorrectly verified, it means that the data in the memory has been completely damaged, and the software cannot continue to maintain it. Then perform step R; if the backup index summary is also verified If it is correct, execute step P; Step P: Obtain the latest storage pointer of the backup index content from the backup index summary, and use the latest storage pointer to read the latest piece of index information in the backup index content; Step Q: extract the latest storage pointer of the data content from the index information in step P, and execute step G; Step R: re-establish storage information, and start from step A again. 2. the method for storing real-time data according to claim 1, is characterized in that, the data of described real-time collection each sensor also comprises: Step A1: The real-time collection includes: channel number, year, month, day, hour, minute, second, data value information, check code data to generate an index and data content on a daily basis; Step A2: Compare the index information including the year, month, day, start address, end address, and check code with the stored index to determine the storage address of the data content; Step A3: Determine whether the stored data needs to be overwritten according to the storage address of the data content. If overwriting is required, delete the content in the overwritten storage area, store the data content in this area, and change the index information; if not, directly Store the data content in the storage address obtained by step A2; Step A4: Determine whether the index information needs to be updated according to the storage situation of the data content, if necessary, change the stored index to new index information, and if not, keep the original index unchanged. 3. the method for storing real-time data as claimed in claim 2, is characterized in that: described storage data content coverage adopts the circular coverage storage mode of circular sequential storage, and the steps are as follows: Step A31: sequentially store in the SD memory card according to the physical address sequence; Step A32: When storing to the maximum physical address of the SD memory card, delete the data of the earliest day in the SD memory card; Step A33: Then continue to store data at the physical address where the stored content was stored on the earliest day to form a circular storage mode, and make corresponding changes to the index content. 4. The method for storing real-time data as claimed in claim 1, wherein said index further comprises: The index includes the physical storage address pointing to the data content of the day, and the storage index adopts the storage mode of the main part and the backup of the double index, and the main part index and the backup index with the same content are stored in different physical addresses in the ferroelectric memory, Each contains verification information; The storage index consists of an index summary and index content. The content of the index summary includes the starting and ending physical addresses of the index content, the total number of storage records, and the remaining space of the physical storage. The index content is the index information pointing to the stored data generated in units of days. 5. the method for storing real-time data as claimed in claim 1, is characterized in that: described check code R (X) polynomial is CRC16 cyclic check code, and its calculation is: $\frac{\mathrm{<span\; class="notranslate">\; B</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&CenterDot;</span>{\mathrm{<span\; class="notranslate">\; 2</span>}}^{\mathrm{<span\; class="notranslate">\; 16</span>}}}{\mathrm{<span\; class="notranslate">\; G</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; Q</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; G</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span>}$ Wherein B(X) is the n-bit binary sequence number to be checked, G(X)=X ¹⁶ +X ¹⁵ +X ² +1 is a fixed polynomial, Q(X) is an integer polynomial, and the remainder of R(X) is Cyclic check code, X is a polynomial free variable. 6. the method for storing real-time data as claimed in claim 1, is characterized in that: described real-time data value is a single-precision floating-point number, represented by 4 bytes, start address and end address are integer numbers, by 4 bytes representation. 7. The method for storing real-time data according to claim 3, characterized in that: the storage capacity of the SD memory card is 256MB˜4GB. 8. The method for storing real-time data according to claim 4, characterized in that: the storage capacity of the ferroelectric memory is 16KB˜128KB. 9. The method for storing real-time data as claimed in claim 1, characterized in that: the embedded collector is a device for transplanting an embedded operating system constituted by an ARM processor, and the software of the device does not transplant the commonly used File system. 10. the method that the method for storing real-time data as claimed in claim 1 is applied on the embedded ARM collector.

Images