CN110602723A - Two-stage bidirectional prediction data acquisition method based on underwater edge equipment - Google Patents
Two-stage bidirectional prediction data acquisition method based on underwater edge equipment Download PDFInfo
- Publication number
- CN110602723A CN110602723A CN201910795830.0A CN201910795830A CN110602723A CN 110602723 A CN110602723 A CN 110602723A CN 201910795830 A CN201910795830 A CN 201910795830A CN 110602723 A CN110602723 A CN 110602723A
- Authority
- CN
- China
- Prior art keywords
- value
- auv
- data
- prediction
- predicted
- 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.)
- Granted
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B13/00—Transmission systems characterised by the medium used for transmission, not provided for in groups H04B3/00 - H04B11/00
- H04B13/02—Transmission systems in which the medium consists of the earth or a large mass of water thereon, e.g. earth telegraphy
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
- H04W24/06—Testing, supervising or monitoring using simulated traffic
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Arrangements For Transmission Of Measured Signals (AREA)
Abstract
A two-stage bidirectional prediction data acquisition method based on underwater edge equipment is characterized in that a bottom layer-edge-cloud three-layer data collection mode is constructed by transferring calculation from a centralized cloud layer to a distributed edge layer. The data are analyzed and predicted based on historical information by utilizing the mobility and the computing power of the underwater edge device, so that the acoustic wave communication is effectively reduced on the premise of ensuring the accuracy of data transmission, and the energy consumption of the sensor is reduced. And in consideration of the differentiation capability of heterogeneous equipment, different prediction algorithms are adopted for two layers of nodes, namely an edge insulating layer and a bottom layer, so that the underwater data collection process is converted into a two-stage bidirectional data prediction process. The invention can be applied to an underwater sensor network data collection mechanism.
Description
Technical Field
The invention relates to the field of data collection in an underwater sensor network, in particular to a two-stage bidirectional prediction data acquisition method based on underwater edge equipment.
Background
In recent years, underwater wireless sensor networks are widely applied to marine energy exploration, environmental index detection, military monitoring, marine disaster event monitoring and the like, so that the underwater wireless sensor networks are widely concerned at home and abroad. The technology of underwater exploration and other applications is advanced, and the detected data types are increasingly complicated, such as underwater temperature, salinity, oxygen content, pH value and CO2P, etc., the data values collected are lengthy and complex. The cost of battery replacement or charging of the ocean bottom sensor node is high, the underwater sound wave transmission rate is low (generally less than or equal to 100kbps), and the underwater sound wave transmission rate is gradually reduced to zero along with the increase of the distance. When data collected by the sensor nodes in unit time needs to be delivered to the sink nodes, the data needs to pass through a multi-hop route, a large amount of energy consumption is consumed, network congestion can be caused, problems such as delay are caused, and the life cycle of the network is reduced. Therefore, how to deliver data to the cloud in time in the underwater data collection process, and minimize the network energy consumption as much as possible is an urgent problem to be solved.
In an underwater sensor network, a traditional data acquisition mode includes Multi-Hop routing (Multi-Hop) or AUV-assisted Multi-Hop routing (AUV-assisted Multi-Hop) based. However, many researchers now propose different data collection methods, such as monitoring perceptually collected data, predicting perceptually collected data, and the like. In the traditional collection method, member nodes associate the gateway nodes with the shortest path, and transmit data packets to AUV (autonomous Underwater vehicle) or next hop nodes through the gateway nodes. In the sensory data acquisition technology (SDA) based on event coverage detection, a single sensor node is likely to reflect errors or abnormalities on equipment or a system, and in consideration of the fact, the sensory data of the sensor nodes are set to be acquired and routed to the sink node only when the occurrence of possible events is detected by adjacent sensor nodes in a collective mode. The strategy avoids unnecessary routing of sensing data which is possibly unfavorable for event detection, reduces energy consumption and increases network capacity. However, such prediction is only used for determining an emergency, which can save energy for the network and cannot be used for data transmission at each time point. Based on a linear prediction method data collection technology (DBP), characteristics are focused on processing new data. The core idea of the method is that a simple model is used for capturing data change trends, an elastic rule is used for calculating and processing existing interference at the same time, and a user can change a parameter control model on the premise that historical data collected by a sensor node is used as a sample, for example, short-term or long-term historical data is selected for training. However, the prediction accuracy of DBP is low, and when the data is in a short-term linear law, the algorithm is suitable, so that it has limitations.
In general, the above techniques suffer from the following drawbacks. Firstly, a common prediction model is generally applied to a land wireless sensor network, and researchers combining a prediction algorithm with an underwater sensor network are few, so that a proper model cannot be analyzed and a proper prediction algorithm cannot be selected according to the particularity of an underwater environment and the characteristics of an acquisition object. Secondly, aiming at the problems that the underwater nodes are weak in computing capacity, small in stored energy and incapable of timely delivering data to cloud layers, no better solution is provided at present.
Disclosure of Invention
The invention mainly aims to overcome the defects in the prior art, provides a two-stage bidirectional prediction data acquisition method based on underwater edge equipment, and ensures the accuracy of a predicted value.
The invention adopts the following technical scheme:
a two-stage bidirectional prediction data acquisition method based on underwater edge equipment is characterized by comprising the following steps:
1) the sensor node at the bottom layer collects data and sends a transmission request signal to the AUV;
2) establishing a first-stage bidirectional prediction-exponential smooth prediction between the AUV and a bottom sensor node; the AUV and the bottom sensor node respectively select and predict the number of times of exponential smoothing according to whether seasonal or trend components exist in the data of the time sequence, after the prediction is finished, the bottom sensor compares the obtained predicted value with the actual acquisition value and sends the comparison result to the AUV, and the AUV corrects the predicted value which is not in line with the expectation according to the comparison result, updates the correct data value and obtains the data collected by prediction;
3) the AUV sends a transmission request signal to an edge layer, and a second-stage bidirectional prediction-ARMA prediction is established between the edge layer and the AUV; the edge layer and the AUV respectively use Kalman filtering to estimate parameters to obtain a predicted value, the AUV compares the predicted value with the data predicted and collected in the step 2) and sends the comparison result to the edge layer, and the edge layer corrects the predicted value which is not in line with expectation according to the comparison result, updates the correct data value and obtains the final data predicted and collected;
4) the edge layer transmits the final predicted and collected data obtained in the step 3) to the cloud layer.
Preferably, when the energy of the bottom-layer sensor node is greater than sigma, and the observed value of the time series is free of seasonal or trend components, a primary exponential smoothing is adopted, and a primary exponential smoothing formula and model are as follows:
Vt (1)is the first exponential smoothing value of the t period, alpha is the weighting coefficient, sigma is the minimum value of the residual energy, v1,v2,...,vtIn order to be a time series of observations,is a predicted value of the t-th period,is the predicted value of the T + T period.
Preferably, when the data variation of the time series presents a straight line trend, a quadratic exponential smoothing is adopted, and the calculation formula and the model are respectively as follows:
Vt (2)is the second exponential smoothing value of the T period, T is the period number from the current period number T to the prediction period,is a predicted value of the T + T period, at、btAre an undetermined model component.
Preferably, if the time-series data variation shows a curve trend, a cubic exponential smoothing method is required, and the calculation formula and the prediction model of the cubic exponential smoothing method are as follows:
wherein, Vt (3)The cubic exponential smoothing value t of the t period is the current period number, and the parameter at、bt、ctIs an undetermined model component, which can be expressed as:
at＝3Vt (1)-3Vt (2)+Vt (3)
preferably, the second-stage bidirectional prediction-ARMA prediction adopts an AR model, an MA model or an ARMA model.
Preferably, in step 2), the bottom sensor compares the obtained predicted value with the actual collected value, and sends the comparison result to the AUV, and the AUV corrects the predicted value that is not expected according to the comparison result, updates the correct data value, and obtains the data collected by prediction, specifically: the bottom sensor calculates the error between the actual acquisition value and the predicted value, compares the error with a preset threshold value, feeds back a value beta to the AUV, and when the value beta is 1, the predicted value accords with the expectation, and the corresponding predicted value in the AUV does not need to be corrected; if beta is 0, the corresponding predicted value in the AUV needs to be corrected, the AUV moves to the bottom layer sensor node, and the correct acquisition value is received by using sound wave communication.
Preferably, in step 3), if the predicted value is not within the preset error range, the AUV moves to the sink node of the edge layer to transmit a correct data value.
As can be seen from the above description of the present invention, compared with the prior art, the present invention has the following advantages:
1. by adopting the method of the invention, the life cycle of the bottom sensor can be prolonged under the condition of limited energy of the nodes.
2. The method introduces edge calculation, responds timely, fully utilizes the calculation capacity of the edge layer, and reduces the bandwidth pressure and cloud layer calculation pressure.
3. The method provided by the invention adopts a two-stage bidirectional structure, so that the accuracy of data transmission is ensured, and the huge energy consumption caused by sound wave communication is reduced.
Drawings
FIG. 1 is a schematic diagram of two-stage predictive collection of underwater data in accordance with the present invention;
FIG. 2 is a block diagram of the edge computing based underwater data collection of the present invention;
FIG. 3 is a graph of the effect of the fit between the first-level predicted values and the actual values in 2016;
FIG. 4 is a fitting effect graph of the first-level predicted values and the actual values in 2017;
FIG. 5 is a fitting effect graph of the first-level predicted values and actual values in 2018;
FIG. 6 is a graph of the effect of two-level predicted and collected value fits (without correction by the mobile node);
FIG. 7 is a graph of the effect of fitting two levels of predicted values to collected values (error corrected in time);
the invention is described in further detail below with reference to the figures and specific examples.
Detailed Description
The invention is further described below by means of specific embodiments.
Edge computing plays an important role in reducing cloud computing burden, so some computing tasks can be shifted from centralized clouds to the edge end. A data acquisition structure of a bottom layer, an edge and a cloud is built, in order to meet the calculation capacity of different underwater nodes, a two-stage prediction algorithm is built by using mobile nodes of an edge layer, namely future data are predicted by using an underwater robot AUV (autonomous Underwater vehicle) based on historical information, so that underwater acoustic communication and huge energy consumption are reduced. Meanwhile, a bidirectional prediction method is adopted, and the accuracy of the predicted value is ensured by simultaneously performing prediction by a transmission party and a collection party each time and enabling a transmission party node to contain a collected value and a predicted value.
Referring to fig. 1, a diagram of an underwater two-stage bidirectional prediction data structure is shown, and a diagram on the left side of the diagram is a simplified diagram of underwater data transmission, and includes an AUV, a common sensor node and a sink node, and shows a two-stage structure, that is: bottom layer-mobile node AUV and mobile node AUV-water surface sink node; the predicted transmission of data is between the transmitting and collecting parties as shown on the right, using historical data, and if the prediction has a large error, an error is indicated, and the AUV moves to the node to receive the correct value. A
Referring to fig. 2, it is a diagram of an edge-based computing underwater data collection architecture (illustrating a structure including a bottom layer, an edge layer and a cloud layer, data of the bottom layer is transmitted to the cloud layer through the edge layer, and the edge layer is closer to the bottom layer and is moved by an AUV to collect sensor data).
A two-stage bidirectional prediction data acquisition method based on underwater edge equipment comprises the following steps:
1) and the sensor node at the bottom layer collects data and sends a transmission request signal to the AUV, and the AUV establish connection communication. The initial conditions of the step comprise a fully charged AUV, a sensor node initial position, an error control range thd and a residual energy minimum value sigma; when the AUV electric quantity is lower than sigma, the AUV moves to the solar charging panel to be charged.
2) And establishing a first-stage bidirectional prediction-exponential smooth prediction between the AUV and the bottom layer sensor node, namely adopting the same first-stage bidirectional prediction-exponential smooth prediction for the AUV and the bottom layer sensor node. And the AUV and the bottom layer sensor node respectively select the times of exponential smoothing according to whether seasonal or trend components exist in the data of the time sequence, and perform prediction. After the prediction is finished, the bottom sensor compares the obtained predicted value with the actual acquisition value, the comparison result is sent to the AUV, and the AUV corrects the predicted value which is not in accordance with the comparison result and updates the correct data value. Then, in the AUV, the correct predicted values and updated data values constitute the data collected by the AUV prediction.
Specifically, a first-stage bidirectional prediction-exponential smoothing prediction model equation is established. As shown in fig. 1, the lower half is a schematic diagram of the first-level bi-directional prediction process, the left half is the motion trajectory of the AUV, and the right half is the prediction process.
3) The AUV sends a transmission request signal to an edge layer, and a second-stage bidirectional prediction-ARMA prediction is established between the edge layer and the AUV; the edge layer and the AUV respectively use Kalman filtering to estimate parameters to obtain predicted values, the AUV compares the predicted values with the data collected in the prediction in the step 2) and sends the comparison result to the edge layer, the edge layer corrects the predicted values which are not expected according to the comparison result, and updates correct data values, so that the correct predicted values and the updated data values form the final data collected in the prediction and obtained by the edge layer.
4) The edge layer transmits the final predicted and collected data obtained in the step 3) to the cloud layer.
In step 1), when the bottom sensor node is energized>σ, time series observation v1,v2,...,vtIn the absence of seasonal or trending components, a one-time exponential smoothing is employed to correct for omissions that may be generated due to hysteresis in creating the smoothing equation. When node njWhen the time series has no obvious trend variation, the observed value of the t +1 th period can be predicted by using the t th period to perform exponential smoothing once. Setting the observed value of the time series as v1,v2,...,vtThen, the first exponential smoothing formula and the model are,
Vt (1)is an exponential smoothing value of the t-th cycle, alpha (0)<α<1) Is a weighting coefficient, σ is a minimum value of residual energy, v1,v2,...,vtIn order to be a time series of observations,is a predicted value of the t-th period,is the predicted value of the T + T period.
When the change of the time series presents a linear trend, the obvious lag deviation is predicted by a primary exponential smoothing method, and the problem is solved by adopting secondary exponential smoothing. The calculation formula and the model are as follows:
Vt (2)is the second exponential smoothing value of the T period, T is the period number from the current period number T to the prediction period,is a predicted value of the T + T period, at、btAre an undetermined model component.
If the time series changes in a curve trend, a cubic exponential smoothing method is required. The calculation formula and the prediction model of the cubic exponential smoothing method are as follows:
wherein, Vt (3)The third exponential smoothing value of the t period is obtained, and t is the current period number; t is the period number from the current period number T to the prediction period;the predicted value of the T + T period is obtained; parameter at、bt、ctAre an undetermined model component, which may be expressed as,
at＝3Vt (1)-3Vt (2)+Vt (3)
the bottom sensor node and AUV are respectivelyExecuting the first stage of prediction process and obtaining the predicted values respectively, e.g. after the bottom sensor node runs the adaptive exponential smoothing algorithm, obtaining the predicted valuesAt this time the bottom sensor node njHas collection value Collect and prediction valueAnd comparing the error between the two with a preset threshold thd, feeding back a value beta to the AUV, wherein when the value beta is 1, the predicted value of the AUV is in an acceptable range, the bottom-layer sensor node does not need to forward a data packet by using sound waves, and when the value beta is 0, the corresponding predicted value in the AUV needs to be corrected, and the AUV moves to the bottom-layer sensor node and receives a correct acquisition value by using sound wave communication.
In step 3), a second-stage bidirectional prediction-ARMA prediction model equation is established, namely, AUV and the edge layer are subjected to second-stage prediction, and corresponding prediction values are obtained respectively. The process is shown in the upper half of fig. 1, the upper left half represents the motion trajectory and the cycle, and the upper right half represents the second-stage prediction process.
The second-level data transmission is transmission between the mobile edge node AUV and the sink node, and an algorithm with high accuracy is adopted, namely an ARMA (auto regression moving average) model is adopted. It is divided into three major categories of AR model, MA model and ARMA model (auto regression moving average model). If the sequence is a non-stationary sequence, a differential model, namely an ARIMA model, can be used. The ARMA (p, q) model is expressed using mathematical formulas as follows:
when p is the number of autoregressive terms, q is the number of moving average terms, and q is 0, the model is degenerated into an AR (P) model; when p is 0, the model degenerates to ma (q) model, c is a constant term, epsilontRepresenting the error value for the t period.
Predicted value in the above equationFrom p previous observations { Xt-1,Xt-2,...,Xt-pQ error values (e) and (c) are calculatedt-1,εt-2,...,εt-pRepresents a linear combination of. Therefore, the present problem requires that an optimal coefficient set be obtainedAnd a coefficient set { theta }1,θ2,...,θqMakes the mean square error minimum, i.e.
Wherein RMSE is the mean square error, whereinRepresents { Xt-1,Xt-2,...,Xt-pMean of.
Parameters of the authority prediction model can be more effectively estimated by adopting the extended Kalman filter EKF, and the basic EKF state space model is as follows:
Hi＝AiHi-1+Vi
Yi＝CiHi+Wi
in the above formula, H is an n-dimensional system state matrix at the time of i, and A is a state transition matrix which changes along with time; v is a system error matrix at the moment i, W is a measurement error matrix at the moment i, Y represents a measurement vector obtained at the moment i, C is a time-varying output matrix at the moment i, and ViAnd WiIs white noise with mean value of zero and no time correlation, independent of each other, ViIs state noise, WiTo observe the noise, both are white noise without correlation, independent of each other. The combined covariance matrix (noise level) of the two is defined as follows:
E[ViVi T]＝Qi
E[WiWi T]＝Ri
wherein Q, R represents the covariance of the state noise and the observation noise, and for the above system, the extended kalman filter algorithm is as follows:
K(i)＝[AiPiCi T][CiPiCi T+Ri]-1
Hi+1＝AiHi+K(i)[Yi-CiHi]
k (i) represents the Kalman gain value, PiRepresents an update of the error covariance,Ci TRepresenting the transpose of the matrix.
The coefficients in the ARMA model are obtained through the three steps of the formula, the prediction value is solved through parameter estimation, and the coefficients are expressed as a C matrix as follows:
if there is no random input and the measurement is noiseless, then an H matrix can be constructed:
H＝[yi...yi-pεi...εi-q]
{yi...yi-prepresents YiThen in the ith period, when the matrix C and H are determined, the predicted value can be expressed as Yi＝CiHi。
If the predicted result Y isiWithin the error range, the success of the second-stage prediction of data transmission (AUV → sink node of the edge layer) is indicated, that is, the collected data does not need to be transmitted by AUV movement, and the predicted data of the edge layer is taken as the final predicted collected data. If errors occurIf so, the AUV moves to the sink node of the edge layer for correction, and transmits the corresponding predicted and collected data obtained in the step 2). And finally, the data collected by the final prediction is obtained by the edge layer and is transmitted to the cloud.
The invention provides an underwater data collection scheme based on a two-stage bidirectional data prediction algorithm of a bottom layer-edge layer-cloud layer structure, which comprises the following steps: the data collection process is converted into a data prediction process, so that acoustic wave communication is reduced, and transmission energy consumption is reduced; the method is characterized in that prediction is carried out in two stages according to different characteristics of underwater nodes, a bidirectional prediction mode is adopted, the accuracy of data is guaranteed, and the residual energy of the nodes is maximized while the data are accurately transmitted.
According to the method of the present application, the recent data of KEO sites in the noaa (national ocean and atomic administration) database is used, the data at the depth of 400m is used as experimental data, partial data is used as historical data, training prediction is performed, and the fitting results are respectively the training results and partial prediction results of 2016, 2017 and 2018 in the last half year, as shown in fig. 3 to 5. Fig. 6 and 7 are graphs of the fitting effect of two-stage predicted values and collected values, and fig. 6 is a graph of the fitting effect that the AUV cannot correct errors timely when errors occur in the first-stage prediction; fig. 7 is a predicted value fitting effect diagram of the AUV correcting errors in time when errors occur in the first-stage prediction.
The above description is only an embodiment of the present invention, but the design concept of the present invention is not limited thereto, and any insubstantial modifications made by using the design concept should fall within the scope of infringing the present invention.
Claims (7)
1. A two-stage bidirectional prediction data acquisition method based on underwater edge equipment is characterized by comprising the following steps:
1) the sensor node at the bottom layer collects data and sends a transmission request signal to the AUV;
2) establishing a first-stage bidirectional prediction-exponential smooth prediction between the AUV and a bottom sensor node; the AUV and the bottom sensor node respectively select and predict the number of times of exponential smoothing according to whether seasonal or trend components exist in the data of the time sequence, after prediction is finished, the bottom sensor compares an obtained predicted value with an actual acquired value and sends a comparison result to the AUV, and the AUV corrects the predicted value which is not expected according to the comparison result, updates a correct data value and obtains predicted and collected data;
3) the AUV sends a transmission request signal to the edge layer, and second-stage bidirectional prediction-ARMA prediction is established between the edge layer and the AUV; the edge layer and the AUV respectively use Kalman filtering to estimate parameters to obtain predicted values, the AUV compares the predicted values with the data collected in the prediction in the step 2) and sends the comparison result to the edge layer, and the edge layer corrects the predicted values which do not accord with the expectation according to the comparison result, updates correct data values and obtains the final data collected in the prediction;
4) the edge layer transmits the final predicted and collected data obtained in the step 3) to the cloud layer.
2. The two-stage bidirectional prediction data acquisition method based on the underwater edge device as claimed in claim 1, wherein when the node energy of the underlying sensor is > σ, and the observed value of the time series is free of seasonal or trend components, a one-time exponential smoothing is adopted, and a one-time exponential smoothing formula and model are as follows:
Vt (1)is the first exponential smoothing value of the t period, alpha is the weighting coefficient, sigma is the minimum value of the residual energy, v1,v2,...,vtIn order to be a time series of observations,is a predicted value of the t-th period,is the predicted value of the T + T period.
3. The two-stage bidirectional prediction data acquisition method based on the underwater edge device as claimed in claim 1, wherein when the data variation of the time series presents a straight line trend, quadratic exponential smoothing is adopted, and the calculation formula and the model are respectively as follows:
Vt (2)is the second exponential smoothing value of the T period, T is the period number from the current period number T to the prediction period,is a predicted value of the T + T period, at、btAre an undetermined model component.
4. The two-stage bidirectional prediction data acquisition method based on the underwater edge device as claimed in claim 1, wherein if the data variation of the time series shows a curve trend, a cubic exponential smoothing method is required, and a calculation formula and a prediction model of the cubic exponential smoothing method are as follows:
wherein, Vt (3)Cubic exponential smoothing for the t-th cycleThe value t is the current cycle number, parameter at、bt、ctIs an undetermined model component, which can be expressed as:
at＝3Vt (1)-3Vt (2)+Vt (3)
5. the two-stage bidirectional predictive data acquisition method based on an underwater edge device as claimed in claim 1, wherein the second-stage bidirectional predictive-ARMA prediction adopts an AR model, an MA model or an ARMA model.
6. The two-stage bidirectional prediction data acquisition method based on the underwater edge device as claimed in claim 1, wherein in step 2), the bottom sensor compares the obtained predicted value with the actual acquisition value, and sends the comparison result to the AUV, and the AUV corrects the predicted value which is not in accordance with the comparison result, updates the correct data value, and obtains the predicted and collected data, specifically: the bottom sensor calculates the error between the actual acquisition value and the predicted value, compares the error with a preset threshold value, feeds back a value beta to the AUV, and when the value beta is 1, the predicted value accords with the expectation, and the corresponding predicted value in the AUV does not need to be corrected; if beta is 0, the corresponding predicted value in the AUV needs to be corrected, the AUV moves to the bottom layer sensor node, and the correct acquisition value is received by using sound wave communication.
7. The two-stage bidirectional prediction data acquisition method based on the underwater edge device as claimed in claim 1, wherein in the step 3), if the predicted value is not within the preset error range, the sink node that the AUV moves to the edge layer transmits the correct data value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910795830.0A CN110602723B (en) | 2019-08-27 | 2019-08-27 | Two-stage bidirectional prediction data acquisition method based on underwater edge equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910795830.0A CN110602723B (en) | 2019-08-27 | 2019-08-27 | Two-stage bidirectional prediction data acquisition method based on underwater edge equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110602723A true CN110602723A (en) | 2019-12-20 |
| CN110602723B CN110602723B (en) | 2022-05-03 |
Family
ID=68855774
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910795830.0A Active CN110602723B (en) | 2019-08-27 | 2019-08-27 | Two-stage bidirectional prediction data acquisition method based on underwater edge equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110602723B (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111137405A (en) * | 2020-01-08 | 2020-05-12 | 中国电子科技集团公司电子科学研究院 | Data transmission system |
| CN111417178A (en) * | 2020-03-17 | 2020-07-14 | 重庆大学 | Industrial internet energy-saving system and method based on machine learning and edge calculation |
| CN111651312A (en) * | 2020-07-28 | 2020-09-11 | 南京宏声科技有限公司 | Three-level bidirectional early warning method and system for equipment fault based on edge calculation |
| CN111866111A (en) * | 2020-07-13 | 2020-10-30 | 郑州轻工业大学 | Edge deployment method for sensing user demand growth trend |
| CN112148722A (en) * | 2020-10-14 | 2020-12-29 | 四川长虹电器股份有限公司 | Monitoring data abnormity identification and processing method and system |
| CN112909928A (en) * | 2021-01-29 | 2021-06-04 | 华侨大学 | Route planning and charging scheduling algorithm for mobile charging vehicle with wireless sensor network |
| CN113411366A (en) * | 2020-03-17 | 2021-09-17 | 中移(上海)信息通信科技有限公司 | Internet of things data linkage method, device, equipment and medium based on edge calculation |
| CN114051983A (en) * | 2021-11-10 | 2022-02-18 | 山东省海洋资源与环境研究院(山东省海洋环境监测中心、山东省水产品质量检验中心) | Intelligent stichopus japonicus breeding monitoring and collecting system and implementation method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103033186A (en) * | 2012-12-30 | 2013-04-10 | 东南大学 | High-precision integrated navigation positioning method for underwater glider |
| US20150208464A1 (en) * | 2014-01-23 | 2015-07-23 | Electronics And Telecommunications Research Institute | Sensor network system and method for processing sensor data |
| CN108683468A (en) * | 2018-04-27 | 2018-10-19 | 河海大学常州校区 | AUV mobile data collection algorithms in underwater sensing network based on data prediction |
-
2019
- 2019-08-27 CN CN201910795830.0A patent/CN110602723B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103033186A (en) * | 2012-12-30 | 2013-04-10 | 东南大学 | High-precision integrated navigation positioning method for underwater glider |
| US20150208464A1 (en) * | 2014-01-23 | 2015-07-23 | Electronics And Telecommunications Research Institute | Sensor network system and method for processing sensor data |
| CN108683468A (en) * | 2018-04-27 | 2018-10-19 | 河海大学常州校区 | AUV mobile data collection algorithms in underwater sensing network based on data prediction |
Non-Patent Citations (2)
| Title |
|---|
| ZHANGBING ZHOU: "Energy-Efficient Event Determination in Underwater WSNs Leveraging Practical Data Prediction", 《IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS》 * |
| 叶惠荣: "基于SNMP与Aglets的分布式水声无线传感器网络管理系统技术研究与设计", 《中国优秀硕士论文全文数据库(信息科技辑)》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111137405A (en) * | 2020-01-08 | 2020-05-12 | 中国电子科技集团公司电子科学研究院 | Data transmission system |
| CN111417178A (en) * | 2020-03-17 | 2020-07-14 | 重庆大学 | Industrial internet energy-saving system and method based on machine learning and edge calculation |
| CN113411366A (en) * | 2020-03-17 | 2021-09-17 | 中移(上海)信息通信科技有限公司 | Internet of things data linkage method, device, equipment and medium based on edge calculation |
| CN111417178B (en) * | 2020-03-17 | 2021-12-31 | 重庆大学 | Industrial internet energy-saving system and method based on machine learning and edge calculation |
| CN111866111A (en) * | 2020-07-13 | 2020-10-30 | 郑州轻工业大学 | Edge deployment method for sensing user demand growth trend |
| CN111651312A (en) * | 2020-07-28 | 2020-09-11 | 南京宏声科技有限公司 | Three-level bidirectional early warning method and system for equipment fault based on edge calculation |
| CN112148722A (en) * | 2020-10-14 | 2020-12-29 | 四川长虹电器股份有限公司 | Monitoring data abnormity identification and processing method and system |
| CN112148722B (en) * | 2020-10-14 | 2022-06-03 | 四川长虹电器股份有限公司 | Monitoring data abnormity identification and processing method and system |
| CN112909928A (en) * | 2021-01-29 | 2021-06-04 | 华侨大学 | Route planning and charging scheduling algorithm for mobile charging vehicle with wireless sensor network |
| CN114051983A (en) * | 2021-11-10 | 2022-02-18 | 山东省海洋资源与环境研究院(山东省海洋环境监测中心、山东省水产品质量检验中心) | Intelligent stichopus japonicus breeding monitoring and collecting system and implementation method |
| CN114051983B (en) * | 2021-11-10 | 2022-08-30 | 山东省海洋资源与环境研究院(山东省海洋环境监测中心、山东省水产品质量检验中心) | Intelligent stichopus japonicus breeding monitoring and collecting system and implementation method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110602723B (en) | 2022-05-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110602723B (en) | Two-stage bidirectional prediction data acquisition method based on underwater edge equipment | |
| Gatsis et al. | Optimal power management in wireless control systems | |
| You et al. | Minimum data rate for mean square stabilizability of linear systems with Markovian packet losses | |
| CN101458325B (en) | Wireless sensor network tracking method based on self-adapting prediction | |
| CN113537514B (en) | Digital twinning-based federal learning framework with high energy efficiency | |
| CN102316496A (en) | Data merging method based on Kalman filtering in wireless sensor network | |
| CN103595568B (en) | A kind of the Internet based on LS-SVM real-Time Signal Transfer method | |
| CN114422349B (en) | Cloud-edge-end-collaboration-based deep learning model training and reasoning architecture deployment method | |
| WO2019205528A1 (en) | Antenna data-based algorithm for predicting solar energy collection power of wireless sensor network node | |
| CN115173923B (en) | Low-orbit satellite network energy efficiency perception route optimization method and system | |
| CN116599602B (en) | Low-energy-consumption long-distance sonar communication system for ocean monitoring | |
| CN108871324A (en) | A kind of underwater passive integrated navigation system decaying adaptive information fusion method | |
| Maestre et al. | Distributed receding horizon Kalman filter | |
| Zhang et al. | Trajectory prediction based on AIS and BP neural network | |
| Mehrnia et al. | Extreme value theory based rate selection for ultra-reliable communications | |
| CN109413746B (en) | Optimized energy distribution method in communication system powered by hybrid energy | |
| CN111565430A (en) | Marine ship wireless network routing method based on predicted track | |
| CN116669089A (en) | Predicting quality of service via channel aggregation | |
| Nassar et al. | Heterogeneous data reduction in WSN: Application to Smart Grids | |
| Liu et al. | Wavelet neural network based link quality prediction for fluctuating low power wireless links | |
| CN112636719B (en) | ILC system input signal filtering method under data loss and channel noise interference | |
| CN115587713A (en) | Marine ranch disaster decision method based on reinforcement learning | |
| CN115622603A (en) | Age minimization optimization method for auxiliary transmission information | |
| Yang et al. | Learning optimal stochastic sensor scheduling for remote estimation with channel capacity constraint | |
| Zhu et al. | Automatic cruise system for water quality monitoring |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |