CN115132189A - Long connection communication connection pool management method based on self-adaptive control strategy - Google Patents
Long connection communication connection pool management method based on self-adaptive control strategy Download PDFInfo
- Publication number
- CN115132189A CN115132189A CN202210656953.8A CN202210656953A CN115132189A CN 115132189 A CN115132189 A CN 115132189A CN 202210656953 A CN202210656953 A CN 202210656953A CN 115132189 A CN115132189 A CN 115132189A
- Authority
- CN
- China
- Prior art keywords
- connection
- time
- long
- pool
- communication connection
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
- 238000004891 communication Methods 0.000 title claims abstract description 84
- 238000007726 management method Methods 0.000 title claims abstract description 15
- 238000011217 control strategy Methods 0.000 title claims abstract description 14
- 238000000034 method Methods 0.000 claims abstract description 27
- 230000003044 adaptive effect Effects 0.000 claims abstract description 22
- 230000006855 networking Effects 0.000 claims description 6
- 239000000126 substance Substances 0.000 claims description 3
- 230000003993 interaction Effects 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 230000004044 response Effects 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000007812 deficiency Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/08—Speech classification or search
- G10L15/18—Speech classification or search using natural language modelling
- G10L15/1822—Parsing for meaning understanding
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/26—Speech to text systems
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/30—Services specially adapted for particular environments, situations or purposes
- H04W4/40—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/30—Services specially adapted for particular environments, situations or purposes
- H04W4/40—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P]
- H04W4/48—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P] for in-vehicle communication
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Artificial Intelligence (AREA)
- Telephonic Communication Services (AREA)
Abstract
The application discloses a long connection communication connection pool management method based on an adaptive control strategy. The long connection communication connection pool management method based on the self-adaptive control strategy comprises the following steps: the first end obtains the connection request number of the second end at a first time point; the method comprises the steps that a first end obtains the average time consumed for completing connection of tasks in a long connection communication connection pool within a preset time period before a first time point; the first end obtains the minimum idle connection number which needs to be kept in the long connection communication connection pool under the condition of a first time point according to the average time consumed by completing the connection of the tasks in the long connection communication connection pool and the connection request number; and adjusting the number of idle connections in the long connection pool according to the minimum number of idle connections. According to the method and the device, the number of the idle connections in the long connection pool is dynamically adjusted according to the minimum number of the idle connections, so that connection availability can be ensured to be available when the connection request of each second end comes, and the time consumption of service requests is reduced.
Description
Technical Field
The application relates to the technical field of vehicle-mounted voice processing, in particular to a long connection communication connection pool management method based on an adaptive control strategy and a vehicle networking system.
Background
Referring to fig. 1, on a vehicle, a driver or passenger completes a voice request in general through three steps:
firstly, a client initiates a request, establishes long connection with an intelligent voice access gateway, and sends an audio request through the long connection;
the voice gateway forwards the audio request requested by the client to a voice recognition service based on long connection, and recognizes an intermediate recognition text result and a final recognition text result corresponding to the voice, the intermediate recognition text result is returned to the client, and the final recognition text result is used for semantic understanding processing;
and thirdly, the voice gateway forwards the final text recognized by the voice to a semantic understanding service to obtain a semantic analysis result, and the result is returned to the client.
In the interaction process, because the voice recognition service realizes the streaming recognition based on the long connection, for the voice recognition service, each request access has network and memory overhead, and in order to save resource overhead and increase throughput, the voice recognition service actively disconnects the long connection after one voice recognition is finished.
In addition, the voice recognition service can also detect whether the connection is idle in a set time period, and if no data flows in and out, the connection can also be actively disconnected, so that the resource overhead is saved.
On the other hand, for the intelligent voice access gateway, in order to implement streaming voice recognition, a new long connection needs to be created to complete voice recognition each time a voice recognition service is requested, so that performance loss caused by frequent connection creation is relatively large, and currently, by testing a streaming voice recognition model service which is mainstream in the industry, average connection time is about 50 ms. This value also rises to between 100ms and 200ms at high concurrency requests. The time consumption caused by establishing the connection can reduce the throughput of voice interaction on one hand and can also cause the service response speed to be obviously reduced.
In the prior art, two normal forms of speech recognition processing are mainly used in the industry at present, one is streaming speech recognition, which refers to a type of ASR model that can support real-time return of recognition results in the process of processing an audio stream. In contrast to the non-streaming model, which must process the complete sentence audio before the result can be returned. The request model has many disadvantages, mainly including the following:
firstly, a client sends a complete voice request at one time, and the failure rate is higher under the condition of poor mobile network environment, which influences the user experience
Secondly, when the voice recognition is carried out, the complete user request audio data is required to complete the recognition, the performance is poor, and the space is not improved.
In the current intelligent voice interaction system under various scenes (such as a vehicle-mounted scene, an intelligent sound box scene, a mobile phone voice assistant scene and the like), in order to improve user experience and interaction response speed, a Streaming voice recognition Model (Streaming ASR Model) has been adopted. In the streaming recognition mode, the long connection is recognized by establishing ASR at each request, and the following problems exist:
firstly, the time consumed for establishing connection each time, the time consumed for overlaying service requests and the response speed are influenced.
Secondly, with the increase of the concurrency, the connection time consumption is increased along with the increase of the system overhead, so that the system response speed is greatly influenced, and the throughput of the system is reduced.
In the existing technology in the industry, there are mature and reliable schemes for short connection scenarios and connection reusable scenarios, such as a database connection pool, an HTTP connection pool, etc., and such conventional ideas are to establish communication connections of a basic number at a time to reduce operations for frequently establishing connections. In addition, such implementations are based on an important premise: the connection can be reused. As mentioned above, for a speech recognition scenario, after a connection is used, a speech recognition service is actively disconnected, or if no data flows in and out within a set time period after the connection is established, the speech recognition service is also actively disconnected due to an idle timeout mechanism, so that there is no mature and reliable technical solution for handling such connection pool management of non-reusable long connections.
Accordingly, a solution is desired to solve or at least mitigate the above-mentioned deficiencies of the prior art.
Disclosure of Invention
The present invention aims to solve at least one of the above problems by providing a method for managing a pool of long-connection communication connections based on an adaptive control strategy.
In one aspect of the present invention, a method for managing a long connection communication connection pool based on an adaptive control policy is provided, where the method for managing a long connection communication connection pool based on an adaptive control policy includes:
the first end obtains the number of connection requests of the second end at a first time point, wherein a request sent by the second end is used for enabling the first end to borrow an idle connection in a long connection pool of the first end and a third end;
the first end acquires the average time consumed for completing connection of tasks in the long connection communication connection pool within a preset time period before a first time point;
the first end obtains the minimum idle connection number which needs to be kept in the long connection communication connection pool under the condition of the first time point according to the average time consumed by completing the connection of the tasks in the long connection communication connection pool and the connection request number;
and the first end adjusts the number of idle connections in the long connection pool according to the minimum number of idle connections.
Optionally, the number of the second ends is multiple.
Optionally, the obtaining, by the first end, an average time taken for completing the connection by the task in the long connection communication connection pool within a preset time period before the first time point includes:
the method comprises the steps that a first end obtains an average value of a connection time-consuming sequence established with a long connection communication connection pool in a preset time period before a first time point;
the method comprises the steps that a first end obtains an average value of a time-consuming sequence of a voice recognition request obtained through a long connection communication connection pool in a preset time period before a first time point;
the first end obtains the average value of the semantic request time-consuming sequence in a preset time period before the first time point;
and according to the average value of the connection time-consuming sequences established with the long connection communication connection pool, acquiring the average value of the voice recognition request time-consuming sequences through the long connection communication connection pool and acquiring the average time-consuming time through the average value of the semantic request time-consuming sequences.
Optionally, the obtaining, by the first end, the minimum number of idle connections that the long connection communication connection pool needs to maintain according to the average time consumed for completing the connection by the task in the long connection communication connection pool and the number of connection requests includes:
the first end obtains the connection request number of the client in a preset time period before the first time point;
and acquiring the minimum idle connection number required to be kept by the connection pool according to the average time consumption and the connection request number.
Optionally, the minimum number of idle connections is calculated using the following formula:
newMinIdle＝[n/(t ave +t’ ave +t nlu )](ii) a Wherein the content of the first and second substances,
t ave establishing an average value of a connection time-consuming sequence for the long-connection communication connection pool; t' ave Acquiring an average value of time-consuming sequences of voice recognition requests for the long connection communication connection pool; t is t nlu is An average of the semantic request time-consuming sequences; the time unit is second; []Is a rounding operation.
Optionally, the obtaining, by the first end, an average value of a connection time-consuming sequence created with the long connection communication connection pool within a preset time period before the first time point includes:
acquiring the time required by the establishment of each connection and the number of the established connections;
and acquiring the average value of the time sequence for establishing the connection with the long connection communication connection pool in a preset time period according to the time required by establishing each connection and the number of the established connections.
Optionally, the obtaining an average value of the time-consuming sequence of the speech recognition requests obtained through the long-connection communication connection pool in a preset time period before the first time point includes:
acquiring the time required by each voice recognition request and the number of created voice recognition requests;
and acquiring the average value of the time-consuming sequences of the voice recognition requests acquired by the long-connection communication connection pool according to the time required by each voice recognition request and the number of the created voice recognition requests.
Optionally, the obtaining an average value of a semantic request time-consuming sequence in a preset time period before the first time point includes:
acquiring the time required by each semantic request and the number of created semantic requests;
and acquiring the average value of the semantic request time-consuming sequences according to the time required by each semantic request and the number of the created semantic requests.
Optionally, the first end is a server, the second end is a client, and the third end is a speech recognition server.
The application also provides a car networking system, car networking system includes: the system comprises a server, a voice recognition server and at least one vehicle, wherein each vehicle comprises a client;
the server, the voice analysis server and the client are matched to realize the long-connection communication connection pool management method based on the adaptive control strategy.
Advantageous effects
The method for managing the long connection communication connection pool based on the adaptive control strategy obtains the minimum number of idle connections based on the average time consumed by completing the connection of tasks in the long connection communication connection pool and the number of connection requests required to request the connection at the current time point, and dynamically adjusts the number of the idle connections in the long connection pool according to the minimum number of the idle connections, so that connection availability can be ensured to be available when the connection request of each second end comes to the greatest extent, and the time consumed by service requests is reduced.
Drawings
Fig. 1 is a data flow framework diagram of a full-link intelligent voice interactive system in the prior art.
Fig. 2 is a flowchart illustrating a method for managing a long connection communication connection pool based on an adaptive control policy according to an embodiment of the present application.
Fig. 3 is a schematic diagram illustrating a configuration of a long connection communication connection pool for a voice recognition request.
Detailed Description
In order to make the implementation objects, technical solutions and advantages of the present application clearer, the technical solutions in the embodiments of the present application will be described in more detail below with reference to the drawings in the embodiments of the present application. In the drawings, the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The described embodiments are a subset of the embodiments in the present application and not all embodiments in the present application. The embodiments described below with reference to the accompanying drawings are illustrative and intended to explain the present application and should not be construed as limiting the present application. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application. Embodiments of the present application will be described in detail below with reference to the accompanying drawings.
Fig. 2 is a flowchart illustrating a method for managing a long connection communication connection pool based on an adaptive control policy according to an embodiment of the present application.
The method for managing the long-connection communication connection pool based on the adaptive control strategy as shown in fig. 2 comprises the following steps:
step 1: the method comprises the steps that a first end obtains the number of connection requests of a second end at a first time point, wherein a request sent by the second end is used for enabling the first end to borrow one idle connection in a long connection pool of the first end and a long connection pool of a third end;
step 2: the first end acquires the average time consumed for completing connection of tasks in the long connection communication connection pool within a preset time period before a first time point;
and step 3: the first end obtains the minimum idle connection number which needs to be kept in the long connection communication connection pool under the condition of the first time point according to the average time consumed by completing the connection of the tasks in the long connection communication connection pool and the connection request number;
and 4, step 4: and the first end adjusts the number of idle connections in the long connection pool according to the minimum number of idle connections.
The method for managing the long connection communication connection pool based on the adaptive control strategy obtains the minimum number of idle connections based on the average time consumed by completing the connection of tasks in the long connection communication connection pool and the number of connection requests required to request the connection at the current time point, and dynamically adjusts the number of the idle connections in the long connection pool according to the minimum number of the idle connections, so that connection availability can be ensured to be available when the connection request of each second end comes to the greatest extent, and the time consumed by service requests is reduced.
In this embodiment, the number of the second ends is plural.
In this embodiment, the acquiring, by the first end, an average time consumed for completing the connection by the task in the long connection communication connection pool within a preset time period before the first time point includes:
the method comprises the steps that a first end obtains an average value of a connection time-consuming sequence established with a long connection communication connection pool in a preset time period before a first time point;
the method comprises the steps that a first end obtains an average value of a time-consuming sequence of a voice recognition request obtained through a long connection communication connection pool in a preset time period before a first time point;
the first end obtains the average value of the semantic request time-consuming sequence in a preset time period before the first time point;
and according to the average value of the connection time-consuming sequences established with the long connection communication connection pool, acquiring the average value of the voice recognition request time-consuming sequences through the long connection communication connection pool and acquiring the average time-consuming time through the average value of the semantic request time-consuming sequences.
In this embodiment, the obtaining, by the first end, the minimum number of idle connections to be maintained by the long connection communication connection pool according to the average time consumed for completing the connection by the task in the long connection communication connection pool and the number of connection requests includes:
the first end obtains the connection request number of the client in a preset time period before the first time point;
and acquiring the minimum idle connection number required to be kept by the connection pool according to the average time consumption and the connection request number.
In this embodiment, the following formula is used for calculation:
newMinIdle＝[n/(t ave +t’ ave +t nlu )](ii) a Wherein the content of the first and second substances,
t ave establishing an average value of a connection time-consuming sequence for the long-connection communication connection pool; t' ave Acquiring an average value of time-consuming sequences of voice recognition requests for the long-connection communication connection pool; t is t nlu is a Mean value of the semantic request time-consuming sequence; the time unit is second; []Is a rounding operation.
In this embodiment, the obtaining, by the first end, an average value of a connection time-consuming sequence established with the long connection communication connection pool within a preset time period before the first time point includes:
acquiring the time required by the establishment of each connection and the number of the established connections;
and acquiring the average value of the time sequence for establishing the connection with the long connection communication connection pool in a preset time period according to the time required by establishing each connection and the number of the established connections.
In this embodiment, the obtaining an average value of the time-consuming sequence of the speech recognition requests obtained through the long-connection communication connection pool in a preset time period before the first time point includes:
acquiring the time required by each voice recognition request and the number of created voice recognition requests;
and acquiring the average value of the time-consuming sequences of the voice recognition requests acquired by the long-connection communication connection pool according to the time required by each voice recognition request and the number of the created voice recognition requests.
In this embodiment, the obtaining an average value of the semantic request time-consuming sequence in a preset time period before the first time point includes:
acquiring the time required by each semantic request and the number of created semantic requests;
and acquiring the average value of the semantic request time-consuming sequences according to the time required by each semantic request and the number of the created semantic requests.
In this embodiment, the first end is a server, the second end is a client, and the third end is a speech recognition server.
The present application is further described in detail below by way of examples, it being understood that the examples do not constitute any limitation to the present application.
When the first end and the third end of this application worked, be provided with long connection communication connection pond between first end and the third end, at first can the initial voice recognition request long connection communication connection pond parameter that sets up: the basic connection pool parameters are initially set according to the configuration file, as shown in fig. 3.
When the method of the present application is performed, first, the first end obtains data: that is, the first end obtains the connection request number n of the client of the intelligent voice access gateway service at the current time point (i.e. at the first time point) from the log.
Acquiring the time required by the creation of each connection and the number of created connections before the current time point (namely within a preset time period before the first time point), and generating p long connection time-consuming sequences for creating the voice recognition according to the time required by the creation of each connection and the number of created connections (t1, t2,.. tp);
acquiring the time required by each voice recognition request and the number of created voice recognition requests before the current time point (namely, within a preset time period before the first time point), and generating time-consuming sequences (t ' 1, t ' 2,.., t ' q) of q voice recognition requests according to the time required by the voice recognition requests and the number of created voice recognition requests;
the time required by each semantic request and the number of created semantic requests before the current time point (i.e. in a preset time period before the first time point) are also acquired, and r time-consuming sequences (t "1, t" 2., t "r) of semantic identification requests are generated according to the time required by each semantic request and the number of created semantic requests.
The average value of the connection time-consuming sequence created by the long-connection communication connection pool in the preset time period before the first time point, the average value of the voice recognition request time-consuming sequence obtained by the long-connection communication connection pool in the preset time period before the first time point, and the average value of the semantic request time-consuming sequence in the preset time period before the first time point are respectively obtained as follows:
calculate the average of p sequences of time spent creating the join:
t ave ＝sum(t 1 ,t 2 ,...,t p )/p
calculate the average of the time-consuming sequence of q speech recognition requests:
t’ ave ＝sum(t’ 1 ,t’ 2 ,...,t’ q )/q
calculating the average value of r time-consuming sequences of semantic requests:
t nlu ＝sum(t” 1 ,t” 2 ,...,t” r )/r
will be described above as t ave、 t’ ave And t nlu The addition is the average elapsed time.
Calculating the minimum number of idle connections to be maintained in the connection pool
newMinIdle＝[n/(t ave +t’ ave +t nlu )]. Wherein [ 2 ], [ 2 ]]Is a rounding operation.
The application also provides a connection pool inactive connection release algorithm: and the connection is closed depending on the idle overtime or the use completion of the ASR service connection, and the inactive connection is cleaned regularly according to the attribute of whether the connection is active, so that the long connection objects in the pool are all available connections.
For example, the first terminal acquires the number of long connections connected with the third terminal in real time, monitors whether the long connections are active, and if one or more of the long connections are inactive, clears the long connections to prevent the situations that some long connections are connected but not used or the system does not disconnect the long connections which should be disconnected.
For example:
suppose that in a long connection pool, there are 10 existing connections, 10 idle connections, 10 preset idle connections, and 50 maximum connections.
At time a, there are 5 connection requests from the second end, and at this time, at time a, there are 15 existing connections and 5 idle connections in the long connection pool.
Assuming that the average time taken for completing the connection by the task in the long connection pool within 10 seconds before the time point a is calculated is obtained by the above formula, the minimum number of idle connections may also be obtained by the above formula.
Assuming that the minimum idle connection number is 10 after the calculation is completed, at this time, it can be seen from the above description that after the time point a, the idle connection number becomes 5, however, the required minimum idle connection number is 10, and 5 is less than 10, and the idle connection number needs to be increased, and in the case that the maximum connection number is not exceeded by 50, the idle connection number may be increased at will according to the needs of the user, for example, 5 may be increased to keep the preset idle connection number always 10, or more may be increased, for example, the number may be increased more than the preset idle connection number, and 10 is increased, and at this time, the idle connection number becomes 15.
By the method, the following advantages can be brought:
1. the minimum available long connection number in the connection pool is judged by combining a plurality of key parameters such as intelligent voice access gateway load, time consumption for establishing connection, time consumption for voice recognition request and the like, connection availability is ensured to be available when the connection request of each client comes as far as possible, and time consumption for establishing connection is saved.
2. Based on the characteristic that the long connection is not reusable when the voice recognition is requested, the realization of the long connection pool is optimized, namely, the inactive connections are periodically expelled, and each connection lent from the pool is ensured to be an available connection.
3. The dynamic balance of the intelligent voice client request connection and the voice recognition request connection is realized based on the long connection pool, so that the service requirement is met, and unnecessary resource occupation is not caused to the voice recognition service.
The application also provides a car networking system, car networking system includes: the system comprises a server, a voice recognition server and at least one vehicle, wherein each vehicle comprises a client;
the server, the voice analysis server and the client are matched to realize the long-connection communication connection pool management method based on the adaptive control strategy.
Although the invention has been described in detail hereinabove with respect to a general description and specific embodiments thereof, it will be apparent to those skilled in the art that modifications or improvements may be made thereto based on the invention. Accordingly, such modifications and improvements are intended to be within the scope of the invention as claimed.
Claims (10)
1. A management method of a long connection communication connection pool based on an adaptive control strategy is characterized in that the management method of the long connection communication connection pool based on the adaptive control strategy comprises the following steps:
the method comprises the steps that a first end obtains the number of connection requests of a second end at a first time point, wherein a request sent by the second end is used for enabling the first end to borrow one idle connection in a long connection pool of the first end and a long connection pool of a third end;
the first end acquires the average time consumed for completing connection of tasks in the long connection communication connection pool within a preset time period before a first time point;
the first end obtains the minimum idle connection number which needs to be kept in the long connection communication connection pool under the condition of the first time point according to the average time consumed by completing the connection of the tasks in the long connection communication connection pool and the connection request number;
and the first end adjusts the number of idle connections in the long connection pool according to the minimum number of idle connections.
2. The method for managing a pool of long-connection communication connections based on an adaptive control strategy according to claim 1, wherein the number of the second ends is plural.
3. The adaptive control policy-based long connection communication connection pool management method according to claim 2, wherein the acquiring, by the first end, an average time taken for a task in the long connection communication connection pool to complete a connection within a preset time period before the first time point includes:
the method comprises the steps that a first end obtains an average value of a connection time-consuming sequence established with a long connection communication connection pool in a preset time period before a first time point;
the method comprises the steps that a first end obtains an average value of a time-consuming sequence of a voice recognition request obtained through a long connection communication connection pool in a preset time period before a first time point;
the first end obtains the average value of the semantic request time-consuming sequence in a preset time period before the first time point;
and according to the average value of the connection time-consuming sequences established with the long connection communication connection pool, acquiring the average value of the voice recognition request time-consuming sequences through the long connection communication connection pool and acquiring the average time-consuming time through the average value of the semantic request time-consuming sequences.
4. The method for managing the long connection communication connection pool based on the adaptive control policy according to claim 3, wherein the obtaining, by the first end, the minimum number of idle connections to be maintained by the long connection communication connection pool according to the average elapsed time for completing the connection by the task in the long connection communication connection pool and the number of connection requests comprises:
the first end obtains the connection request number of the client in a preset time period before the first time point;
and acquiring the minimum idle connection number required to be kept by the connection pool according to the average time consumption and the connection request number.
5. The adaptive control policy-based long connection communication connection pool management method according to claim 4, wherein the minimum number of idle connections is calculated using the following formula:
newMinIdle＝[n/(t ave +t’ ave +t nlu )](ii) a Wherein the content of the first and second substances,
t ave establishing an average value of a connection time-consuming sequence for the long-connection communication connection pool; t' ave Acquiring an average value of time-consuming sequences of voice recognition requests for the long connection communication connection pool; t is t nlu is An average of the semantic request time-consuming sequences; the time unit is second; []Is a rounding operation.
6. The method as claimed in claim 5, wherein the step of the first end obtaining an average value of a sequence of connection times spent in creating the connection with the long-connection communication connection pool within a preset time period before the first time point comprises:
acquiring the time required by the establishment of each connection and the number of the established connections;
and acquiring the average value of the time sequence for establishing the connection with the long connection communication connection pool in a preset time period according to the time required by establishing each connection and the number of the established connections.
7. The adaptive control policy-based long-connection communication connection pool management method according to claim 5, wherein the obtaining an average value of a time-consuming sequence of obtaining the voice recognition request through the long-connection communication connection pool within a preset time period before the first time point comprises:
acquiring the time required by each voice recognition request and the number of created voice recognition requests;
and acquiring the average value of the time-consuming sequences of the voice recognition requests acquired by the long-connection communication connection pool according to the time required by each voice recognition request and the number of the created voice recognition requests.
8. The adaptive control policy-based long connection communication connection pool management method according to claim 5, wherein the obtaining an average value of the semantic request time-consuming sequence within a preset time period before the first time point comprises:
acquiring the time required by each semantic request and the number of created semantic requests;
and acquiring the average value of the semantic request time-consuming sequences according to the time required by each semantic request and the number of the created semantic requests.
9. The adaptive control policy-based long connection communication connection pool management method according to any one of claims 1 to 8, wherein the first end is a server, the second end is a client, and the third end is a voice recognition server.
10. A vehicle networking system, comprising: the system comprises a server, a voice recognition server and at least one vehicle, wherein each vehicle comprises a client;
the server, the voice parsing server and the client cooperatively implement the method for managing the long connection communication connection pool based on the adaptive control policy according to any one of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210656953.8A CN115132189A (en) | 2022-06-10 | 2022-06-10 | Long connection communication connection pool management method based on self-adaptive control strategy |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210656953.8A CN115132189A (en) | 2022-06-10 | 2022-06-10 | Long connection communication connection pool management method based on self-adaptive control strategy |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115132189A true CN115132189A (en) | 2022-09-30 |
Family
ID=83378975
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210656953.8A Pending CN115132189A (en) | 2022-06-10 | 2022-06-10 | Long connection communication connection pool management method based on self-adaptive control strategy |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115132189A (en) |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050172029A1 (en) * | 2004-01-29 | 2005-08-04 | International Business Machines Corporation | Method and apparatus for managing a connection pool using heuristic information |
| CN105785929A (en) * | 2014-12-22 | 2016-07-20 | Tcl集团股份有限公司 | WeChat-based home electric appliance control method and system |
| CN108075915A (en) * | 2016-11-16 | 2018-05-25 | 西北工业大学 | A kind of RDMA communication connection pond management methods based on ADAPTIVE CONTROL |
| CN108924145A (en) * | 2018-07-16 | 2018-11-30 | 百度在线网络技术（北京）有限公司 | Network transfer method, device and equipment |
| US20190005047A1 (en) * | 2017-06-30 | 2019-01-03 | Paypal, Inc. | Intelligent database connection management |
| CN110289993A (en) * | 2019-06-06 | 2019-09-27 | 阿里巴巴集团控股有限公司 | A kind of connection pool management method, apparatus, equipment and adapting system |
| CN111477229A (en) * | 2020-04-15 | 2020-07-31 | 苏州思必驰信息科技有限公司 | Voice recognition request processing method and device |
| CN111586140A (en) * | 2020-04-30 | 2020-08-25 | 深圳市优必选科技股份有限公司 | Data interaction method and server |
| CN111770157A (en) * | 2020-06-24 | 2020-10-13 | 广东浪潮大数据研究有限公司 | Business processing method and device, electronic equipment and storage medium |
| CN111818667A (en) * | 2020-09-10 | 2020-10-23 | 深圳传音控股股份有限公司 | Device connection method, device and readable storage medium |
| CN112217857A (en) * | 2019-07-12 | 2021-01-12 | 电子湾有限公司 | Connection service discovery and load rebalancing |
| CN112732441A (en) * | 2021-01-07 | 2021-04-30 | 山东众阳健康科技集团有限公司 | Connection method, device and system of multi-element dynamic connection pool |
| CN114286425A (en) * | 2021-12-17 | 2022-04-05 | 北京达佳互联信息技术有限公司 | Long connection establishing system, method and device, electronic equipment and storage medium |
-
2022
- 2022-06-10 CN CN202210656953.8A patent/CN115132189A/en active Pending
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050172029A1 (en) * | 2004-01-29 | 2005-08-04 | International Business Machines Corporation | Method and apparatus for managing a connection pool using heuristic information |
| CN105785929A (en) * | 2014-12-22 | 2016-07-20 | Tcl集团股份有限公司 | WeChat-based home electric appliance control method and system |
| CN108075915A (en) * | 2016-11-16 | 2018-05-25 | 西北工业大学 | A kind of RDMA communication connection pond management methods based on ADAPTIVE CONTROL |
| US20190005047A1 (en) * | 2017-06-30 | 2019-01-03 | Paypal, Inc. | Intelligent database connection management |
| CN108924145A (en) * | 2018-07-16 | 2018-11-30 | 百度在线网络技术（北京）有限公司 | Network transfer method, device and equipment |
| CN110289993A (en) * | 2019-06-06 | 2019-09-27 | 阿里巴巴集团控股有限公司 | A kind of connection pool management method, apparatus, equipment and adapting system |
| CN112217857A (en) * | 2019-07-12 | 2021-01-12 | 电子湾有限公司 | Connection service discovery and load rebalancing |
| CN111477229A (en) * | 2020-04-15 | 2020-07-31 | 苏州思必驰信息科技有限公司 | Voice recognition request processing method and device |
| CN111586140A (en) * | 2020-04-30 | 2020-08-25 | 深圳市优必选科技股份有限公司 | Data interaction method and server |
| CN111770157A (en) * | 2020-06-24 | 2020-10-13 | 广东浪潮大数据研究有限公司 | Business processing method and device, electronic equipment and storage medium |
| CN111818667A (en) * | 2020-09-10 | 2020-10-23 | 深圳传音控股股份有限公司 | Device connection method, device and readable storage medium |
| CN112732441A (en) * | 2021-01-07 | 2021-04-30 | 山东众阳健康科技集团有限公司 | Connection method, device and system of multi-element dynamic connection pool |
| CN114286425A (en) * | 2021-12-17 | 2022-04-05 | 北京达佳互联信息技术有限公司 | Long connection establishing system, method and device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2021088964A1 (en) | Inference system, inference method, electronic device and computer storage medium | |
| CN110557451A (en) | Dialogue interaction processing method and device, electronic equipment and storage medium | |
| CN110839084A (en) | Session management method, device, equipment and medium | |
| CN107690149B (en) | Method for triggering network policy update, management function entity and core network equipment | |
| CN108650667A (en) | Terminal scheduling method and apparatus | |
| CN111638948A (en) | Multi-channel high-availability big data real-time decision making system and decision making method | |
| CN111225451A (en) | Halt processing method, system, MME and computer readable storage medium | |
| CN109167737A (en) | The CAN dispatching method of fused data reduction and dynamic priority queue | |
| CN110868323B (en) | Bandwidth control method, device, equipment and medium | |
| CN113012695B (en) | Intelligent control method and device, electronic equipment and computer readable storage medium | |
| CN112820295B (en) | Voice processing device and system, cloud server and vehicle | |
| CN112104867B (en) | Video processing method, video processing device, intelligent equipment and storage medium | |
| CN115132189A (en) | Long connection communication connection pool management method based on self-adaptive control strategy | |
| CN106341433A (en) | Service system, signal server and intercom server | |
| CN111422078B (en) | Electric vehicle charging data allocation monitoring method based on block chain | |
| CN110392019B (en) | Method and system for establishing special load bearing of cluster audio and video service | |
| CN113873475B (en) | Vehicle-mounted interconnected audio data processing method and device and electronic equipment | |
| CN111901401B (en) | Method and device for balancing SMF load of session management function | |
| CN116415973A (en) | Model training method, customer service agent matching method, server and storage medium | |
| CN114500237A (en) | Communication method and system | |
| CN113225512A (en) | Audio and video distribution method and system for seat terminal of video call center | |
| CN113055538B (en) | Emergency rescue method for vehicle and related product | |
| CN111740956B (en) | Vehicle communication method, device, computer readable medium and electronic equipment | |
| CN112367634B (en) | Message sending method and device | |
| CN107612747B (en) | Calling service scheduling method and system based on cloud |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |