CN111726407B - Fog calculation monitoring technology for famous flowers and medicinal plants cultivation in intelligent plant factory environment - Google Patents
Fog calculation monitoring technology for famous flowers and medicinal plants cultivation in intelligent plant factory environment Download PDFInfo
- Publication number
- CN111726407B CN111726407B CN202010552281.7A CN202010552281A CN111726407B CN 111726407 B CN111726407 B CN 111726407B CN 202010552281 A CN202010552281 A CN 202010552281A CN 111726407 B CN111726407 B CN 111726407B
- Authority
- CN
- China
- Prior art keywords
- data
- monitoring
- model
- sensing data
- fog
- 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.)
- Active
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 85
- 238000004364 calculation method Methods 0.000 title claims abstract description 28
- 238000005516 engineering process Methods 0.000 title claims abstract description 27
- 238000005070 sampling Methods 0.000 claims abstract description 74
- 238000001914 filtration Methods 0.000 claims abstract description 35
- 238000000034 method Methods 0.000 claims abstract description 32
- 230000003044 adaptive effect Effects 0.000 claims abstract description 30
- 230000005540 biological transmission Effects 0.000 claims abstract description 6
- 230000008859 change Effects 0.000 claims description 33
- 230000001953 sensory effect Effects 0.000 claims description 14
- 238000003860 storage Methods 0.000 claims description 14
- 230000008635 plant growth Effects 0.000 claims description 11
- 230000007423 decrease Effects 0.000 claims description 8
- 230000000694 effects Effects 0.000 claims description 8
- 238000012806 monitoring device Methods 0.000 claims description 6
- 238000012800 visualization Methods 0.000 claims description 6
- 230000001052 transient effect Effects 0.000 claims description 5
- 230000000007 visual effect Effects 0.000 claims description 5
- 238000001514 detection method Methods 0.000 claims description 4
- 230000000737 periodic effect Effects 0.000 claims description 3
- 230000035945 sensitivity Effects 0.000 claims description 3
- 230000003247 decreasing effect Effects 0.000 claims description 2
- 238000009499 grossing Methods 0.000 claims 1
- 230000008569 process Effects 0.000 abstract description 15
- 230000007613 environmental effect Effects 0.000 abstract description 9
- 230000012010 growth Effects 0.000 abstract description 4
- 238000012545 processing Methods 0.000 abstract description 4
- 230000009286 beneficial effect Effects 0.000 abstract description 3
- 241000196324 Embryophyta Species 0.000 description 62
- CURLTUGMZLYLDI-UHFFFAOYSA-N Carbon dioxide Chemical compound O=C=O CURLTUGMZLYLDI-UHFFFAOYSA-N 0.000 description 14
- 229910002092 carbon dioxide Inorganic materials 0.000 description 7
- 239000001569 carbon dioxide Substances 0.000 description 7
- 230000004044 response Effects 0.000 description 5
- 238000009826 distribution Methods 0.000 description 4
- 239000002689 soil Substances 0.000 description 4
- 238000005286 illumination Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 239000006185 dispersion Substances 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 235000015097 nutrients Nutrition 0.000 description 2
- 241000894006 Bacteria Species 0.000 description 1
- 241000238631 Hexapoda Species 0.000 description 1
- 241000607479 Yersinia pestis Species 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 230000003749 cleanliness Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000003912 environmental pollution Methods 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 230000029553 photosynthesis Effects 0.000 description 1
- 238000010672 photosynthesis Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 230000008929 regeneration Effects 0.000 description 1
- 238000011069 regeneration method Methods 0.000 description 1
- 238000013179 statistical model Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 230000002459 sustained effect Effects 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 230000001131 transforming effect Effects 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/02—Protocols based on web technology, e.g. hypertext transfer protocol [HTTP]
- H04L67/025—Protocols based on web technology, e.g. hypertext transfer protocol [HTTP] for remote control or remote monitoring of applications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Y—INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT]
- G16Y10/00—Economic sectors
- G16Y10/05—Agriculture
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Y—INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT]
- G16Y20/00—Information sensed or collected by the things
- G16Y20/10—Information sensed or collected by the things relating to the environment, e.g. temperature; relating to location
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Y—INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT]
- G16Y40/00—IoT characterised by the purpose of the information processing
- G16Y40/20—Analytics; Diagnosis
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/12—Protocols specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks
-
- 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
- H04W4/38—Services specially adapted for particular environments, situations or purposes for collecting sensor information
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/80—Services using short range communication, e.g. near-field communication [NFC], radio-frequency identification [RFID] or low energy communication
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W84/00—Network topologies
- H04W84/18—Self-organising networks, e.g. ad-hoc networks or sensor networks
Landscapes
- Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Computing Systems (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Environmental & Geological Engineering (AREA)
- Medical Informatics (AREA)
- Biomedical Technology (AREA)
- Toxicology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Agronomy & Crop Science (AREA)
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Economics (AREA)
- General Business, Economics & Management (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention relates to a fog calculation monitoring technology for famous flowers and medicinal plant cultivation in an intelligent plant factory environment, which comprises the following steps: step 1, building a sensing network, carrying out regional division on a plant factory, and uploading acquired sensing data to a fog acquisition module; and 2, after each local device collects the sensing data, estimating the subsequent data scale locally through a lightweight data scale estimation model to adjust the forwarding amount, processing the sensing data through a local adaptive sampling model and an adaptive filtering model, and finally forwarding the processed sensing data to the cloud server through the forwarding device according to the adjusted forwarding amount. The invention has the beneficial effects that: the invention estimates the transmission flow of environmental parameter data collected by a sensor in the growth process of famous flowers based on a lightweight estimation model, filters the sensing data flow by a self-adaptive sampling and self-adaptive filtering technology, and realizes real-time monitoring of appropriate amount of effective sensing data by using fog calculation.
Description
Technical Field
The invention relates to the technical field of monitoring based on a fog calculation mode, in particular to a fog calculation monitoring technology for cultivating famous flowers and medicinal plants in an intelligent plant factory environment; the method comprises the steps of establishing a low-power consumption approximate estimation model, a self-adaptive sampling model and a self-adaptive filtering model for sensing data acquisition in the plant factory environment, and realizing environmental parameter monitoring of famous flowers and medicinal plant cultivation in the intelligent plant factory environment.
Background
With the development of modern biotechnology, planting technology and internet of things technology, plant factories use heat-insulating and lightproof materials as an enclosure structure, strictly control gas exchange with the external environment, and adopt artificial light sources to provide illumination for plants, so that energy and substance flow in the system forms a self-circulation regeneration system along with air circulation, thereby being beneficial to improving indoor carbon dioxide concentration, promoting photosynthesis to shorten plant cultivation period, and preventing bacteria and insect pests from entering the room to ensure higher product cleanliness. Compared with the traditional artificial cultivation mode, the plant factory fine cultivation is not influenced by the outside and can be precisely controlled, the high-density plant cultivation can be carried out, and the annual stable production of high-quality plants is realized. In addition, the limitation on plant factory site selection is small, production can be carried out in cities according to local conditions, and the problems of environmental pollution and storage caused by automobile transportation in the logistics link are reduced.
The key point of industrial fine cultivation of famous flowers and medicinal plants is to provide special environments required for growth for different growth stages of different plant varieties and accelerate the growth of plants so as to shorten the cultivation period. Therefore, monitoring various environmental object information (such as temperature, illumination, moisture, soil pH value and the like) of the plant factory is a precondition and a core for realizing industrial fine cultivation of famous flowers and medicinal plants. The existing monitoring technology for fine factory cultivation of famous flowers and medicinal plants can be mainly divided into two steps: 1) various environment sensors collect relevant information such as temperature, illumination, moisture, soil pH value and the like according to preset frequency; 2) and transmitting the acquired information to a background monitoring center in a wireless transmission mode such as 4G, WiFi or Zigbee. However, the amount of data generated by various environmental sensors in plant factories is rapidly increasing every day, and plant fine-grained factory cultivation puts higher demands on the response time and safety of environmental information monitoring technology. Although the existing monitoring technology for plant factory fine cultivation provides an efficient computing platform for processing various environment sensing data, the increase speed of the network bandwidth is far from the increase speed of the data, the decrease speed of the network bandwidth cost is much slower than that of hardware resource costs such as a CPU (Central processing Unit) and a memory, and meanwhile, the network delay is hardly improved in a breakthrough manner due to the complex network environment. Therefore, the existing monitoring technology for fine factory cultivation of famous flowers and medicinal plants needs to solve two major bottlenecks of network bandwidth and delay.
Disclosure of Invention
The invention aims to overcome the defects in the prior art and provide a fog calculation monitoring technology for cultivating famous flowers and medicinal plants in an intelligent plant factory environment.
This kind of famous flowers and medicinal plant cultivation factory monitored control system under intelligent plant factory environment includes: the system comprises a plant factory data acquisition module, a fog acquisition module, a cloud storage module and a monitoring system visualization module; the plant factory data acquisition module is connected with the fog acquisition module through WiFi or Bluetooth for data transmission, the fog acquisition module uploads data to the cloud storage module, and the visual module of the monitoring system reads data from the cloud storage module.
Preferably, a sensor is arranged in the plant factory data acquisition module, and the sensor comprises a temperature sensor, a carbon dioxide sensor and a humidity sensor; the fog acquisition module is used for carrying out data scale estimation, self-adaptive sampling and self-adaptive filtering; the cloud storage module is used for uploading data to a cloud end and performing local backup on the data; the monitoring system visualization module is used for visualizing environmental data, plant growth and nutrient solution conditions.
The fog calculation monitoring technology of the monitoring system of the famous flower and medicinal plant cultivation factory in the intelligent plant factory environment comprises the following steps:
step 1, building a sensing network, carrying out regional division on a plant factory, dividing a monitoring region for each local fog computing device, and deploying a plurality of local devices in each region by a plant factory data acquisition module to acquire sensing data such as temperature, humidity and carbon dioxide; uploading acquired sensing data to a fog acquisition module, such as temperature and humidity, carbon dioxide, soil pH value and the like; to be forwarded subsequently;
step 2, after each local device collects the sensing data, the subsequent data scale is estimated locally through a lightweight data scale estimation model to adjust the forwarding amount, meanwhile, the sensing data is processed through a local adaptive sampling model and an adaptive filtering model, and finally, the processed sensing data is forwarded to a cloud server through forwarding devices according to the adjusted forwarding amount;
step 2.1, estimating the subsequent data scale through a probability index weighted moving average model and trend detection according to the scale of the sensing data, updating the probability index weighted moving average model on line and adjusting the data forwarding amount:
deploying a lightweight data scale estimation model on local equipment, and predicting monitoring data of a specific environment monitoring device at a future moment by using the monitoring data of the specific environment monitoring device in the past period; two consecutive data points v i And v i-1 Distance δ between values i The definition is as follows:
δ i ＝|v i -v i-1 |
by distance δ i Updating the change of the local reference running time of the sensing data flow rho (M), and calculating the current sensing data flow change through moving average, wherein the change is expressed as mu i Let the distance between the next two data points be δ i+1 (ii) a The distance between two consecutive values represents a change in the stream of sensory data; approximate prediction of monitoring sensing data is carried out by using a probability exponential weighting moving average model, and a weighting factor (0) is introduced<α<1) The older sensor data is weighted down in an exponential change pattern, as shown in the following equation:
in the above formula, with delta 1 To initializeu 1 Then, updating in an iterative manner; although the exponentially weighted moving average model is more suitable for the monitoring requirements of an actual plant, its response to transient changes is unstable, so it cannot always be assumed that only exponential weighting exists. Specifically, if the probability index weighted moving average model encounters an abrupt peak after a long period of settling phase, and this abrupt peak is followed by a settling phase, the probability index weighted moving average model retains this peak; this will result in overestimating the subsequent δ i Therefore, the accuracy of the approximate data scale estimation model is influenced;
step 2.2, deploying a self-adaptive sampling model facing plant factory refined cultivation monitoring on local equipment, calculating confidence according to the scale of the sensing data, comparing the confidence with inaccuracy defined by a user, adjusting a sampling period and sampling the sensing data; the core of the self-adaptive sampling is to dynamically adjust the sampling periodicity T based on the change of the plant factory monitoring sensing data stream i The monitoring precision still meets the precision requirement given by the user;
in addition, and based only on step functions (e.g., T) i+1 ←T i ±T step ，T step Radius of confidence interval) to adjust the sampling rate; the proposed adaptive algorithm can be based on confidence on the flow of fast-changing metrics in the appropriate range T min ，T max ]Carrying out efficient response internally; step 2.3, deploying a self-adaptive filtering model facing plant factory refined cultivation monitoring on local equipment, and filtering the sampled sensing data: the filter range is denoted R if the value v i ∈[v i-1 -R,v i-1 +R]Then filter v i Current data point d of i (ii) a Dynamically adjusting a filtering range R according to the Fano factor and the change of the monitored sensing data flow, and adjusting a sampling period according to a filtering result;
step 2.4, deploying a self-adaptive forwarding model facing to plant factory refined cultivation monitoring on local equipment, forwarding the forwarding quantity obtained by sampling or filtering data according to data scale estimation model estimation, and transmitting the data to a cloud server;
and 3, the cloud server receives and analyzes the sensing data forwarded by the local equipment, stores different types of sensing data into a local database, displays plant growth condition data or plant growth environment data on the equipment through a visual chart, and provides real-time data monitoring for managers.
Preferably, the local device in step 1 comprises a sensor.
Preferably, in order to solve the problem that the probability index weighted moving average model suffers from a sudden data peak change, the probability index weighted moving average model is solved through two steps of probability weighting and trend detection, and the probability index weighted moving average model is deployed at each local device. Continuously estimating the scale of subsequent data according to the sensing data stream, updating the model on line and adjusting the forwarding amount; the method specifically comprises the following steps:
step 2.1.1, calculating subsequent data volume changes by using a probability index weighting method based on sensor data: the effect of sudden transient changes in the data is accommodated using variable weighting factors:
in the above formula, u 1 For current sense data stream changes, δ i For two successive data points v i And v i-1 The distance between values;
for the weighting factor to be probabilistically variable,
P i is the weight of the ith iteration;
the value is the current delta i The probability of (d); beta is P i The weight of (c); alpha is a weighting factorSeed, 0<α<1; the principle of the probability index weighted moving average model is that the current delta i There is a contribution with probability p to the estimation process. Thus, the weight is updated to 1- β P i So as to limit the model overestimation of the subsequent delta by taking into account unexpected peaks that suddenly have little effect on the subsequent estimation i (ii) a If an unexpected peak is followed by a persistent change in the sensory data stream, subsequent unexpected peaks will be assigned a larger p-value, allowing them to have a greater impact on the estimation process;
step 2.1.2, although the probability exponential weighted moving average model avoids over-estimation of the unexpected peaks by the model, it does not take into account the monotonic phase of the upward and downward trends, which tends to introduce time lag effects in the estimation process. Thus, the model proposed by Holt et al was used to estimate the monotonic increase or monotonic decay in the change in the sensory data stream:
in the above formula, x i Is the current sensing data stream; u. of 1 Is the current sensory data stream change; delta i For two successive data points v i And v i-1 The distance between values; xi is [0,1 ]]A smooth weight in the range, a value of ξ close to 1, represents a preference for recent trends; to initialize a calculation X 1 Taking the initial value of i as 2; hysteresis effects are reduced by raising the moving average of a probability exponential weighted moving average model to an appropriate numerical basis during the estimation process
And 2.1.3, deploying a data scale estimation model at each local device according to the calculation result, estimating the scale of subsequent sensing data according to the sensing data flow, updating the probability index weighted moving average model on line and adjusting the data forwarding amount.
Preferably, the step 2.2 specifically comprises the following steps:
step 2.2.1, according to the current sampling period T i Monitoring the estimated sampling period T of the next data point i+1 : increasing the estimated sampling period T of the next data point if the load decreases i+1 Decreasing the next data point estimated sampling period T if the load increases i+1 (ii) a The next data point estimates the sampling period T i+1 The magnitude of the increase or decrease depends on the confidence c i Representing the confidence level c of the adaptive sampling model estimate and following the present change of the sensed data stream i When sensing data stream confidence c i When the sampling time is larger, the self-adaptive sampling model adopts a larger sampling period;
step 2.2.2, updating the data scale estimation model and calculating the confidence coefficient c of the sensing data flow i This is then compared to a user-defined imprecision γ, γ ∈ [0,1 ]](ii) a While calculating a new sampling period T i+1 ：
In the above formula, λ is a weight coefficient, c i Is the confidence level; γ is imprecision, and if γ → 0, the above equation converges to the periodic sampling method; if γ → 1, each sampling interval is adjusted even if no reliable estimate can be made; thus, if the data size estimation model fails to provide an estimate within some confidence interval, the adaptive sampling model will be at the next data estimation point d i+1 Roll back to default sampling period T min 。
Preferably, the step 2.3 specifically comprises the following steps:
step 2.3.1, calculating the Fano factor over a time window:
in the above formula, σ 2 Is the variance, μ is the mean, σ i And mu i Are all estimated by data sizeCalculating the change probability P provided by the counting model by weighting calculation; the Fano factor is denoted W as the variance σ 2 The ratio to the mean μ;
step 2.3.2, finish F i Then, the error variance σ is calculated err Compared to the maximum inaccuracy y provided by the user: if F i Indicates that the current sensing data stream is not spread and sigma err If the value is less than gamma, the filtering range is enlarged, and nearby values are filtered out, and meanwhile, the data are kept in the precision requirement defined by the user; if F i Indicating that the current sensory data stream is overly dispersed, the filtering range is shortened or restored to a default value and an anomaly in the data is reported.
Preferably, the imprecision parameter γ in step 2.2.2 is used to set the sensitivity.
Preferably, the time complexity of the adaptive sampling method of the adaptive sampling model in step 2.2 is constant time, since all calculations are based on pre-collected values and the entire sensing data stream information is not required.
The invention has the beneficial effects that: the invention estimates the transmission flow of environmental parameter data collected by a sensor in the growth process of famous flowers based on a lightweight estimation model, filters the sensing data flow by the self-adaptive sampling and self-adaptive filtering technology, and realizes the real-time monitoring of proper and effective sensing data by utilizing fog calculation.
Drawings
FIG. 1 is a schematic view of a sensor network topology of a fine cultivation factory for famous flowers and precious medicinal plants;
FIG. 2 is a flow diagram of a plant factory fog calculation monitoring technique;
FIG. 3 is a flow chart of a lightweight approximation sensor flow value estimation model;
fig. 4 is a flow chart of adaptive sampling and adaptive filtering.
Detailed Description
The present invention will be further described with reference to the following examples. The following examples are set forth merely to aid in the understanding of the invention. It should be noted that, for a person skilled in the art, several modifications can be made to the invention without departing from the principle of the invention, and these modifications and modifications also fall within the protection scope of the claims of the present invention.
The invention processes the sensor data of flowers and medicinal plants based on a fog computing mode, and extends and expands the computing, network and storage capacities of the cloud of the current plant factory monitoring system to the network edge.
First, the overall idea of the invention:
the invention mainly solves the following problems: the invention provides a plant factory monitoring technology based on a fog calculation mode, which aims to solve the network bandwidth and delay bottleneck faced by the monitoring of the fine cultivation environmental parameters of famous flowers and medicinal plants in plant factories. On one hand, the technology carries out data processing locally, so that the calculation time of data preprocessing is shorter and the required bandwidth is less, and the requirement of a monitoring system on a network is reduced. The monitoring system local end applying the fog calculation mode needs to selectively screen, sample, filter and forward the sensing data. The sensing data are processed and transmitted to the cloud server, so that the required storage space is smaller, and the data can be monitored and visualized more conveniently.
Second, example
Example 1
The utility model provides a famous flower and medicinal plant cultivation mill monitored control system under intelligence plant factory environment, includes: the system comprises a plant factory data acquisition module, a fog acquisition module, a cloud storage module and a monitoring system visualization module; plant factory data acquisition module carries out data transmission through wiFi or bluetooth connection fog acquisition module, fog acquisition module uploads data to cloud storage module, the visual module of monitoring system reads data from cloud storage module. A sensor is arranged in the plant factory data acquisition module and comprises a temperature sensor, a carbon dioxide sensor and a humidity sensor; the fog acquisition module is used for carrying out data scale estimation, self-adaptive sampling and self-adaptive filtering; the cloud storage module is used for uploading data to a cloud end and performing local backup on the data; the monitoring system visualization module is used for visualizing environmental data, plant growth and nutrient solution conditions.
Example 2
The fog calculation monitoring technology of the famous flower and medicinal plant cultivation factory monitoring system in the intelligent plant factory environment comprises the following steps, as shown in figure 1:
A. the method comprises the following steps of carrying out region division on a plant factory, deploying a plurality of local devices in each region to collect sensing data such as temperature, humidity and carbon dioxide:
the step A comprises the following steps:
and (4) building a sensing network, and collecting related information such as temperature and humidity, carbon dioxide, soil pH value and the like according to a monitoring strategy of the fog calculation module. And dividing a monitoring area for each local fog computing device, and transmitting the related data acquired by all the sensors in the monitoring area to the local device for subsequent forwarding.
B. After each local device collects the sensing data, the subsequent data scale is estimated locally through a lightweight data scale estimation model to adjust the forwarding amount, meanwhile, the sensing data is subjected to adaptive filtering through a local adaptive sampling and filtering model, and finally, the processed sensing data is forwarded to the cloud server through the forwarding device according to the adjusted forwarding amount, as shown in fig. 2:
the step B comprises the following steps:
b11, deploying a lightweight sensing data scale estimation model on the local device, and predicting the monitoring data of the specific environment monitoring device at the future time by using the monitoring data of the specific environment monitoring device in the past time.
Two consecutive data points v i And v i-1 Distance δ between values i The definition is as follows:
δ i ＝|v i -v i-1 |
distance delta i For updating the local reference run-time variation of the sensory data stream ρ (M), the current sensory data stream variation is calculated by moving average, denoted as μ i . Let the distance of the next two data points be δ i+1 . Intuitively, the distance between two successive values represents a change in the sensed data streamAnd (4) transforming. Approximate prediction of monitoring sensing data is carried out by using an exponential weighted moving average model, and a weighting factor (0) is introduced<α<1) The older sensed data is weighted down in an exponential change pattern as shown in the following equation:
in the above formula, with delta 1 To initialize u 1 Followed by an iterative update. Although the exponentially weighted moving average model is more suitable for the monitoring requirements of an actual plant, its response to transient changes is unstable, so it cannot always be assumed that only exponential weighting exists. Specifically, if the exponentially weighted moving average model encounters a sudden peak after a long period of settling phase, and if this sudden peak is followed by a settling phase, the model will retain this peak. This will result in overestimating the subsequent δ i And thus may affect the accuracy of the approximate estimation model.
In order to solve the problem that the model suffers from a sudden data peak change, the model is solved through two steps of probability weighting and trend detection, and the model is deployed at each local device. And continuously estimating the scale of subsequent data according to the sensing data stream, updating the model on line and adjusting the forwarding amount. The step B11 specifically includes the following three steps:
b111, using variable weighting factors to adapt to the influence of sudden transient changes in the data, as follows:
the probability index weighted moving average model introduces a weighting factor that is variable in probability
In the above formula, P i Is the weight of the i-th iteration,
the value is the current delta i Following a modeled distribution of changes in the sensed data stream, β is P i The weight of (c). The principle of the probability index weighted moving average model is that the current delta i There is a contribution to the estimation process with probability p. Thus, the weight is updated to 1- β P i So as to limit the model overestimation of the subsequent δ by considering in the estimation process "unexpected" peaks that are sudden but have little effect on the subsequent estimation i . If an "unexpected" peak is followed by a sustained change in the sensory data stream, subsequent "unexpected" peaks will be assigned larger p-values, allowing them to have a greater impact on the estimation process.
B112, while the probability exponential weighted moving average model avoids over-estimation of the unexpected peaks by the model, it does not account for the monotonic phase of the upward and downward trends, which tends to introduce time lag effects in the estimation process. Thus, the model proposed by Holt et al was used to estimate the monotonic increase/decay in the metric flow variation as shown by:
wherein ξ is [0,1 ]]A smooth weight in the range, with a value close to 1, represents a preference for recent trends. The initial value of i is taken to be 2 to initialize the calculation of X 1 Thus, hysteresis effects are reduced in the estimation process by raising the moving average of the probability exponential weighted moving average model to the appropriate numerical basis
And B113, integrating the two steps, and deploying the model at each local device. And continuously estimating the scale of subsequent data according to the sensing data stream, updating the model on line and adjusting the forwarding amount.
B12, deploying an adaptive sampling model facing plant factory refined cultivation monitoring on local equipment, and carrying out sensing dataThe rows are sampled and the sampling period is adjusted as shown in fig. 3. The core of the self-adaptive sampling is to dynamically adjust the sampling periodicity T based on the change of the plant factory monitoring sensing data stream i While the monitoring accuracy still meets the accuracy requirements given by the user.
Monitoring the estimated sampling period T of the next data point i+1 Dependent on the current sampling period T i The estimated sampling period for the next data point is increased if the load decreases and vice versa. The magnitude of the increase or decrease required depends on the confidence c i The confidence with which the model estimates and follows the current change in the metric flow is represented. When the sensing data flow estimation model compares "confidence" (confidence c) i Larger), adaptive sampling will take a larger sampling period. Hence, and is based only on data point values v i Compared with the threshold sampling technology for adjusting the sampling rate, the self-adaptive sampling method provided by the invention considers the change condition of the sensing data flow and the confidence degree of model estimation at the same time.
After updating the estimation model and calculating the estimation confidence, it is compared to an acceptable user-defined inaccuracy (denoted as γ). Imprecision parameter (gamma ∈ [0,1 ]]) For setting the sensitivity while calculating a new sampling period T i+1 As follows:
in the above equation, λ is a weighting coefficient, and if γ → 0, the above equation will converge to the periodic sampling method. Conversely, if γ → 1, an adjustment occurs every sampling interval even if a reliable estimate cannot be made. Thus, if the sensor stream data estimation model fails to provide an estimate within some confidence interval, the adaptive sampling algorithm will be at the next data estimation point d i+1 Roll back to default sampling period T min . In addition, and based only on step functions (e.g., T) i+1 ←T i ±T step ) Step-by-step technique for adjusting the sampling rate step Is the confidence interval radius; proposed fromThe adaptation algorithm may be based on confidence in the appropriate range T for a rapidly changing stream of metrics min ，T max ]And high-efficiency response is performed.
The time complexity of the proposed adaptive sampling method is constant time, since all calculations are based on pre-collected values and the entire sensing data stream information is not required. The inaccuracy y is the only parameter that is self-defined in the estimation process.
B13, deploying an adaptive filtering model facing plant factory fine cultivation monitoring on the local device, filtering the sampled sensing data, and adjusting the sampling period according to the filtering result, as shown in fig. 4. Plant monitoring sensory data stream filtering is the process of suppressing data points when successive data point values are less than a range, denoted as R. Therefore, if v i ∈[v i-1 -R,v i-1 +R]Then filter the value v i Current data point d of i . However, this requires that the user has a prior known distribution of data point values, and that this distribution does not change, otherwise there is no guarantee that any values will be filtered.
By the adaptive filtering technology, the filtering range R is dynamically adjusted according to the change of the monitored sensing data stream, and meanwhile, the precision requirement defined by a user is still met. The filtering range R depends on the changes in the monitored sensory flow data, where only a small portion of the data needs to be precision limited, and R is adjusted in steps based on the number of previously filtered data points.
The Fano factor is used to display the degree of change associated with monitoring the current change in the sensory data stream. The Fano factor (F is more than or equal to 0) is the same as the dispersion index, is the normalization measurement of the dispersion of probability distribution, and is used for quantifying whether a group of data points are clustered or not compared with a statistical model (F)<1) Or dispersed. The Fano factor is calculated over a time window, denoted W, as the variance σ 2 The ratio to the mean μ is shown by the following formula:
variance σ of the above formula 2 The sum mean μ does not require additional computation because σ i And mu i The data flow estimation method is obtained by weighting calculation of change probability P provided by a sensing data flow estimation model; nor does it use a window of data points, because σ i And mu i The contribution of each data point is adjusted accordingly by weighting according to the previous value. Intuitively, when σ i Reduced, Fano factor F i Following the change, a decrease in the size of the data stream is indicated. F is calculated i Then, will σ err The (error variance) is compared to the maximum tolerable inaccuracy provided by the user, denoted as γ. If F i Indicating that the sensed data stream is not spread and sigma err Less than γ, the filtering range becomes larger, attempting to filter out nearby values while still keeping the data as a whole within the user-defined accuracy requirements. Otherwise, if F i If the current sensing data flow is excessively dispersed, the filtering range is shortened or restored to a default value, and an exception in the data is reported.
As with adaptive sampling, the adaptive filtering algorithm has a temporal and spatial complexity of O (1), since R i+1 Is calculated from its previous value, and mu i ，σ i And σ err Is the output of the runtime estimation model from the foregoing.
B14, deploying an adaptive forwarding model facing plant factory fine cultivation monitoring on local equipment, forwarding the sampled and filtered sensing data according to the forwarding amount estimated by the sensing data scale estimation model, and transmitting the data to the cloud server. With the mode of fog calculation, the forwarding process requires less bandwidth and thus reduces the demand on the network.
C. The cloud server receives and stores the sensing data, displays important data such as plant growth condition data and plant growth environment data on the equipment, and provides real-time data monitoring for management personnel:
the step C comprises the following steps:
and receiving and analyzing the plant factory sensing data forwarded by the local equipment through the cloud server, and storing different types of sensing data into a local database.
And displaying various sensing data information of the plant factory in real time through a proper visual chart.
Thirdly, experimental conclusion:
the system can estimate the subsequent data scale through the data scale acquired by the sensor so as to adjust the forwarding amount, and performs adaptive sampling and filtering according to the current data so as to forward the sensor data with proper scale and low noise to the cloud server. The cloud server can realize real-time display of the plant growth condition data and the plant growth environment data while storing the data. Therefore, under the condition of remote monitoring, the culturist can also realize more accurate understanding of the greenhouse environment, and the method has universality and can be popularized and applied to various crops.
Claims (7)
1. The utility model provides a famous flowers and medicinal plant cultivation factory monitoring system's fog calculation monitoring technology under intelligence plant factory environment, monitored control system includes: the system comprises a plant factory data acquisition module, a fog acquisition module, a cloud storage module and a monitoring system visualization module; the plant factory data acquisition module is connected with the fog acquisition module through WiFi or Bluetooth for data transmission, the fog acquisition module uploads data to the cloud storage module, and the monitoring system visualization module reads the data from the cloud storage module; the method is characterized in that the fog calculation monitoring technology comprises the following steps:
step 1, building a sensing network, carrying out regional division on a plant factory, dividing a monitoring region for each local fog computing device, and deploying a plurality of local devices in each region by a plant factory data acquisition module to acquire sensing data; uploading the acquired sensing data to a fog acquisition module;
step 2, after each local device collects the sensing data, the subsequent data scale is estimated locally through a lightweight data scale estimation model to adjust the forwarding amount, meanwhile, the sensing data is processed through a local adaptive sampling model and an adaptive filtering model, and finally, the processed sensing data is forwarded to a cloud server through forwarding devices according to the adjusted forwarding amount;
step 2.1, estimating the subsequent data scale through a probability index weighted moving average model and trend detection according to the scale of the sensing data, updating the probability index weighted moving average model on line and adjusting the data forwarding amount:
deploying a lightweight data scale estimation model on local equipment, and predicting monitoring data of a specific environment monitoring device at a future moment by using the monitoring data of the specific environment monitoring device in the past period; two consecutive data points v i And v i-1 Distance δ between values i The definition is as follows:
δ i ＝|v i -v i-1 |
by distance delta i Updating the change of the local reference running time of the sensing data flow rho (M), and calculating the current sensing data flow change through moving average, which is expressed as mu i Let the distance between the next two data points be δ i+1 (ii) a Approximate prediction of monitoring sensing data is carried out by using a probability index weighted moving average model, and a weighting factor (0) is introduced<α<1) The sensed data is weighted down in an exponential change pattern, as shown in the following equation:
in the above formula, with delta 1 To initialize u 1 ；
Step 2.2, deploying a self-adaptive sampling model facing plant factory refined cultivation monitoring on local equipment, calculating confidence according to the scale of the sensing data, comparing the confidence with inaccuracy defined by a user, adjusting a sampling period and sampling the sensing data;
step 2.3, deploying a self-adaptive filtering model facing plant factory refined cultivation monitoring on local equipment, and filtering the sampled sensing data: the filter range is denoted R if the value v i ∈[v i-1 -R,v i-1 +R]Then filter v i Current data point d of i (ii) a And dynamically adjusting the filtering range according to the Fano factor and the change of the monitored sensing data streamAdjusting the sampling period according to the filtering result;
step 2.4, deploying an adaptive forwarding model facing plant factory refined cultivation monitoring on local equipment, forwarding the sampled or filtered data according to the forwarding amount obtained by estimation of the data scale estimation model, and transmitting the data to a cloud server;
and 3, the cloud server receives and analyzes the sensing data forwarded by the local equipment, stores different types of sensing data into a local database, displays plant growth condition data or plant growth environment data on the equipment through a visual chart, and provides real-time data monitoring for managers.
2. The fog computing monitoring technology of the famous flower and medicinal plant cultivation factory monitoring system in the intelligent plant factory environment as claimed in claim 1, wherein: the local device in step 1 comprises a sensor.
3. The fog computing monitoring technology of the famous flower and medicinal plant cultivation factory monitoring system in the intelligent plant factory environment as claimed in claim 1, wherein said step 2.1 specifically comprises the following steps:
step 2.1.1, calculating subsequent data volume changes by using a probability index weighting method based on sensor data: the weighting factors are used to accommodate the effects of sudden transient changes in the data:
in the above formula, u 1 For current sense data stream changes, δ i For two successive data points v i And v i-1 The distance between values;
in order to be a weighting factor, the weighting factor,
P i is the weight of the ith iteration;
the value is the current delta i The probability of (d); beta is P i The weight of (c); α is a weighting factor, 0<α<1; updating the weight to 1-beta P i ；
Step 2.1.2, estimating monotonic increase or monotonic decay in the change of the sensory data stream using model proposed by Holt:
in the above formula, x i Is the current sensing data stream; u. of 1 Is the current sensory data stream change; delta i For two successive data points v i And v i-1 The distance between values; xi is [0,1 ]]A smoothing weight within a range; taking the initial value of i as 2;
and 2.1.3, deploying a data scale estimation model at each local device according to the calculation result, estimating the scale of subsequent sensing data according to the sensing data flow, updating the probability index weighted moving average model on line and adjusting the data forwarding amount.
4. The fog computing monitoring technology of the famous flower and medicinal plant cultivation factory monitoring system in the intelligent plant factory environment as claimed in claim 1, wherein said step 2.2 specifically comprises the steps of:
step 2.2.1, according to the current sampling period T i Monitoring the estimated sampling period T of the next data point i+1 : increasing the estimated sampling period T of the next data point if the load decreases i+1 Decreasing the next data point estimated sampling period T if the load increases i+1 (ii) a The next data point estimates the sampling period T i+1 The magnitude of the increase or decrease depends on the confidence c i When sensing data stream confidence c i When larger, the adaptive sampling model is adopted to be largerThe sampling period of (a);
step 2.2.2, updating the data scale estimation model and calculating the confidence coefficient c of the sensing data flow i This is then compared to a user-defined imprecision γ, γ ∈ [0,1 ]](ii) a While calculating a new sampling period T i+1 ：
In the above formula, λ is a weight coefficient, c i Is the confidence level; γ is imprecision, if γ → 0, the above equation will converge to the periodic sampling method; if γ → 1, adjust each sampling interval; if the data size estimation model fails to provide an estimate within some confidence interval, the adaptive sampling model will be at the next data estimation point d i+1 Roll back to default sampling period T min 。
5. The fog computing monitoring technology of the famous flower and medicinal plant cultivation factory monitoring system in the intelligent plant factory environment as claimed in claim 1, wherein said step 2.3 specifically comprises the steps of:
step 2.3.1, calculating the Fano factor over a time window:
in the above formula, σ 2 Is the variance, μ is the mean, σ i And mu i All are obtained by weighting calculation of the change probability P provided by a data scale estimation model; the Fano factor is denoted W as the variance σ 2 The ratio to the mean μ;
step 2.3.2, finish F i Then, the error variance σ is calculated err Compared to the maximum inaccuracy y provided by the user: if F i Indicates that the current sensing data stream is not spread and sigma err If the value is less than gamma, the filtering range is enlarged, and nearby values are filtered; if F i Indicating that the current sensory data stream is overly dispersed, the filtering range is shortened or restored to a default value and an anomaly in the data is reported.
6. The fog calculation monitoring technology of the famous flower and medicinal plant cultivation factory monitoring system in the intelligent plant factory environment as claimed in claim 4, wherein: the imprecision parameter y in said step 2.2.2 is used to set the sensitivity.
7. The fog computing monitoring technology of the famous flower and medicinal plant cultivation factory monitoring system in the intelligent plant factory environment as claimed in claim 1, wherein: the time complexity of the adaptive sampling method of the adaptive sampling model in step 2.2 is constant time.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010552281.7A CN111726407B (en) | 2020-06-17 | 2020-06-17 | Fog calculation monitoring technology for famous flowers and medicinal plants cultivation in intelligent plant factory environment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010552281.7A CN111726407B (en) | 2020-06-17 | 2020-06-17 | Fog calculation monitoring technology for famous flowers and medicinal plants cultivation in intelligent plant factory environment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111726407A CN111726407A (en) | 2020-09-29 |
| CN111726407B true CN111726407B (en) | 2022-08-02 |
Family
ID=72567002
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010552281.7A Active CN111726407B (en) | 2020-06-17 | 2020-06-17 | Fog calculation monitoring technology for famous flowers and medicinal plants cultivation in intelligent plant factory environment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111726407B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112527829B (en) * | 2020-12-17 | 2022-05-10 | 浙江经贸职业技术学院 | Industrial data transmission and visualization system based on Internet of things |
| CN114913029B (en) * | 2022-04-29 | 2023-05-02 | 云铂(宁夏)科技有限公司 | Intelligent agricultural monitoring platform based on Internet of things |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170048308A1 (en) * | 2015-08-13 | 2017-02-16 | Saad Bin Qaisar | System and Apparatus for Network Conscious Edge to Cloud Sensing, Analytics, Actuation and Virtualization |
| CN105610944B (en) * | 2015-12-29 | 2019-03-05 | 北京物联远信息技术有限公司 | A kind of mist computing architecture of internet of things oriented |
| CN106647514A (en) * | 2016-12-28 | 2017-05-10 | 安徽工程大学 | Cement enterprise carbon emission real-time on-line monitoring management system |
| CN107172166B (en) * | 2017-05-27 | 2021-03-23 | 电子科技大学 | Cloud and mist computing system for industrial intelligent service |
| CN108198439B (en) * | 2018-01-24 | 2021-05-14 | 浪潮集团有限公司 | Urban intelligent traffic control method based on fog calculation |
| CN109862011B (en) * | 2019-02-01 | 2021-03-30 | 华南理工大学 | Real-time monitoring system for environment of Internet of things based on fog calculation |
-
2020
- 2020-06-17 CN CN202010552281.7A patent/CN111726407B/en active Active
Non-Patent Citations (1)
| Title |
|---|
| "Maximum Data-Resolution Efficiency for Fog-Computing Supported Spatial Big Data Processing in Disaster Scenarios";JunBo Wang;《 IEEE Transactions on Parallel and Distributed Systems 》;20190130;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111726407A (en) | 2020-09-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Bai et al. | Reliable data fusion of hierarchical wireless sensor networks with asynchronous measurement for greenhouse monitoring | |
| CN112817354A (en) | Livestock and poultry house culture environment temperature prediction control system and regulation and control method thereof | |
| Choudhary et al. | Autonomous crop irrigation system using artificial intelligence | |
| CN111726407B (en) | Fog calculation monitoring technology for famous flowers and medicinal plants cultivation in intelligent plant factory environment | |
| CN110069032B (en) | Eggplant greenhouse environment intelligent detection system based on wavelet neural network | |
| CN110119767B (en) | Intelligent cucumber greenhouse temperature detection device based on LVQ neural network | |
| CN114911831A (en) | Agricultural real-time early warning method and system based on Flink streaming calculation and deep learning | |
| CN118395386B (en) | Plant growth monitoring method for intelligent plant factory | |
| CN114818888B (en) | Soil composition data fusion method and system based on multi-channel Kalman filtering | |
| CN115907366A (en) | Agricultural product growth environment optimal regulation and control method and equipment based on flamingo algorithm | |
| CN114626010A (en) | Irrigation quantity calculation method and system based on Catboost | |
| Ali et al. | Time-series prediction for sensing in smart greenhouses | |
| CN114092776A (en) | Multi-sensor data fusion method applied to intelligent agriculture | |
| CA3196136A1 (en) | Advanced crop manager for crops stress mitigation | |
| US20200019858A1 (en) | Optimization of cyber-physical systems | |
| CN116520917A (en) | Internet of things intelligent optimization monitoring method and system for agricultural product environment of greenhouse | |
| CN116957141A (en) | Intelligent irrigation prediction method and system based on transducer model | |
| Adeyemo | Soft Computing techniques for weather and Climate change studies | |
| Agyeman et al. | LSTM-based model predictive control with discrete inputs for irrigation scheduling | |
| WO2020255678A1 (en) | Information processing device and method | |
| Pahuja et al. | An intelligent wireless sensor and actuator network system for greenhouse microenvironment control and assessment | |
| CN118295254B (en) | Agricultural product cultivation environment control method and device and Internet of things system | |
| JP7437061B2 (en) | Growth environment prediction device, growth environment control system, and growth environment prediction method | |
| KR20200084407A (en) | Neural network based nutrient solution control system and method | |
| CN117704566B (en) | Indoor temperature and humidity control system and method based on new energy wind cabinet data identification |
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 |