CN112056591A

CN112056591A - Method for smoothly and controllably heating rate of coffee beans and automatically identifying roasting state

Info

Publication number: CN112056591A
Application number: CN202010967247.6A
Authority: CN
Inventors: 陈科明; 俞锋; 安建伟; 曾佳
Original assignee: Hangzhou Dianzi University
Current assignee: Hangzhou Dianzi University
Priority date: 2020-09-15
Filing date: 2020-09-15
Publication date: 2020-12-11
Anticipated expiration: 2040-09-15
Also published as: CN112056591B

Abstract

The invention discloses a method for smoothly and controllably heating up coffee beans and automatically identifying the roasting state. The invention designs a dynamic controllable smoothness adjusting mechanism aiming at the requirements of different stages in coffee roasting on smoothness of the heating rate. In order to ensure that the baking state can be identified and updated in time, the parameters of the current bean species are obtained from a rear-end server before baking, and a greedy algorithm is used for a double-layer linked list structure during baking to obtain local optimal smoothness, obtain global optimal smoothness and ensure the timely recording of the baking state transition point. And iteratively updating the baking parameters of various beans by a back-end server through a machine learning feedback mechanism.

Description

Method for smoothly and controllably heating rate of coffee beans and automatically identifying roasting state

Technical Field

The invention relates to the technical field of intelligent coffee machines in the Internet of things, in particular to a method for smoothly and controllably identifying the temperature rise rate and the roasting state of coffee beans.

Background

With the improvement of living standard, coffee has become one of the most popular beverages, and the roasting quality of coffee beans directly determines the taste quality of coffee. Therefore, coffee roasting requires a professional roasting engineer to select coffee beans and to control the time, especially the temperature rise rate. The method for calculating the heating rate of the coffee machine in the market at present has the following problems:

firstly, the method comprises the following steps: since the conventional coffee roasting heating rate refers to a temperature rising within 30 seconds, there is a problem that the heating rate is not accurate. In particular, some stages in baking require higher accuracy, which requires that the temperature rise rate be smoothness controlled and baking state analysis be performed dynamically.

Secondly, the method comprises the following steps: the coffee beans are required to undergo a steaming stage, a dehydration stage, a first explosion stage and a second explosion stage in the roasting process, wherein the temperature return point, the first explosion point, the second explosion point and the like are required to be recorded. The traditional coffee roasting adopts manual recording and has the problem that the temperature return point moves to the right, and the problem that the roasting state is inconvenient to identify and is not strict exists.

Disclosure of Invention

The invention aims to provide a method for smoothly and controllably heating up coffee beans and automatically identifying the roasting state, which solves the problem that the smoothness of the roasting heating up rate of the traditional coffee machine cannot be intelligently regulated and controlled according to the roasting stage and the requirements of customers, and simultaneously solves the problems that the temperature return point is inaccurate and the one-explosion two-explosion point cannot be automatically recorded in the roasting process.

The technical scheme of the invention is as follows:

on one hand, a coffee roasting temperature data storage system is established, and the system comprises an Android hand used by a client

Machine storage and back-end server storage. The coffee machine transmits the temperature of each second to the mobile phone through Socket communication, the mobile phone receives the temperature and stores the temperature in a local Room library and uploads the temperature to a back-end server through the connection of the router, and the server needs to store baking parameters of various beans. And through baking feedback of different bean species by a baker, recording a more accurate baking stage temperature interval and the change condition of the heating rate when each stage is switched, and obtaining the optimal parameters through iterative learning of a neural network algorithm.

The stored data comprises temperature data and smoothness data, and when a user uses a mobile phone to display an image, the temperature data can be processed through the smoothness data, and an expected temperature rise rate curve can be obtained through setting required smoothness.

On the other hand, a method for controlling the smoothness of the temperature rise rate and identifying the baking state is provided.

Firstly, the problem of controllable smoothness of a heating rate is solved, a temperature linked list is created according to a smoothness value x set by a user, the temperature linked list is transmitted into each time a temperature value is received, and the temperature value of a queue head is popped up each time a temperature value is transmitted into the linked list after the length of the linked list reaches x. The heating rate R is equal to the tail end L minus the head end F of the linked list, divided by the length x and multiplied by 60, and the controllability is that the length x of the linked list can be adjusted manually or intelligently.

Secondly, the automatic identification of the baking state is realized, the function of marking on the curve is realized, a heating rate linked list with the fixed length of 8 is established under the temperature linked list, the heating rate of 8 seconds is referred to bean parameters obtained by a rear-end server, and the temperature return point, the first explosion point, the second explosion point and the like are obtained through judgment.

The invention has the beneficial effects that: aiming at the temperature with the heating rate which is not smooth enough and the accuracy of intelligent adjustment cannot be achieved, the invention designs the double-layer linked list to automatically adjust the smoothness of the heating rate at different stages of baking, so that the key process is more accurate, and other processes are smoother. The invention also automatically identifies the change of the state in the coffee roasting process by identifying the change of the heating rate in the heating rate linked list.

Drawings

The invention is further illustrated with reference to the following figures and examples.

FIG. 1 shows an overall frame diagram;

FIG. 2 illustrates the processing and elements of a doubly linked list;

fig. 3 shows a user usage flow and a data processing flow.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be further described in detail with reference to the accompanying drawings and specific embodiments.

Fig. 1 shows an overall framework diagram, and the coffee machine completes Socket communication with the mobile phone through the mobile phone and a distribution network thereof, and communicates with a back-end server through a router. The Socket short-distance communication is used for ensuring that baking data can obtain accurate temperature data in real time in the environment with poor network.

The temperature module of coffee machine body detects the temperature and carries out the preliminary treatment to send the beans body temperature toward back end server and cell-phone end, calculate through beans body temperature and draw the intensification rate, when the temperature module finds that the temperature is unusual, send unusual frame to the cell-phone, provide alarming function.

FIGS. 2 and 3 illustrate a double linked list structure and process flow

When the user starts baking, the user selects whether to start the automatic baking state recognition mode, if not, the smoothness of the heating rate is set to be 15s, and the smooth processing is carried out by taking 15 seconds as a base number on the representative heating rate. Initially there is no data in the temperature linked list, so the formula cannot be R =60 (L-F)/x; when the length x of the linked list does not reach the length set by the user, the curve is not smooth, and particularly when the length is only 2/3, the temperature rise rate may appear in a peak valley shape due to uneven heating of the bean body. In order to solve the problem of uneven heating, the coffee machine measures 5 points per second to carry out temperature preprocessing, and the incorrect temperature is eliminated, so that the sent temperature is the average temperature of the bean bodies. And when x does not reach the set length, judging whether abnormal temperature exists or not during the calculation of the linked list, and then obtaining the correct heating rate.

When the temperature reaches the manual setting length, the heating rate is calculated according to a formula, but the temperature is still the definition of the heating rate of the traditional coffee machine, namely the temperature which is increased within a period of time. In order to obtain a more accurate temperature at that time, the tail end temperature L and the third tail end temperature L3 of the temperature chain table in fig. 2 and the head temperature F are calculated to obtain a heating rate of R =60(2L-F-L3)/(x +3), where x is the smoothness of the setting and x > = 3; the rate of temperature rise is most accurate when x =3, but the smoothness is low, with a peak-valley plot.

If the user does not manually set the smoothness, the temperature profile enters an automatic adjustment state. Before this, bean species identification is performed (parameters of different beans are different) to obtain parameters, and then coffee roasting state identification is performed according to the parameters. The current optimal smoothness x is used after the recognition is finished by using the idea of a greedy algorithm (the bean type parameters at each stage are considered to be correct), and the formula of the temperature rise rate is still R =60(2L-F-L3)/(x + 3). The difference is that x is dynamically changed according to parameters (temperature threshold), and the final global optimal smoothness is achieved through the local optimal smoothness. The characteristics of the water-washed yarrowia beans are discussed below:

coffee roasting is divided into 4 stages: steaming and roasting stage, dewatering stage, first explosion stage and second explosion stage. The specific operation steps comprise preheating the coffee machine, reminding a user to discharge beans after preheating is finished, starting baking at the moment, and receiving temperature data by the mobile phone. Before the temperature return point appears, the temperature of the machine is reduced due to the addition of beans, and the temperature rise rate is negative; therefore, the values in the heating rate linked list are all negative numbers, and when the heating rate is gradually close to zero and the rotation is positive, the temperature is automatically identified and recorded as a temperature return point; this temperature return point should occur at the lowest end of the temperature curve, otherwise the calculation is not accurate.

The occurrence of the temperature return point represents entering into a steam baking mode which is extremely important for temperature control, and at the moment, if the automatic adjustment mode is adopted, the default smoothness is 5, and the high-precision low-smoothness mode is adopted. The subsequent state identification also needs to depend on the identification of temperature, and the robustness of the identification only by the temperature rise rate is not high. The method mainly comprises the steps of bean temperature segmentation, wherein the dehydration stage is 140-180 ℃, the first explosion is 180-210 ℃, and the second explosion is 210-ending. For example, the one shot characteristic of the Jerserphine washes: coffee beans exhibit a significant exothermic reaction, resulting in a decrease in the rate of temperature rise. The accurate time point of the first explosion can be obtained by matching the temperature rise rate linked list with the temperature linked list. Two explosion points and two dehydration points are obtained in the same way, and the smoothness x is reduced when the temperature is close to 140 ℃, 180 ℃ and 210 ℃ in the same way, so that the precision is improved. After identification, x rises, improving the smoothness of the curve.

And finally, the mobile phone stores the recorded data of the temperature return point, the one-shot point and the like in a mobile phone database and a back-end server for reference and inspection of a coffee baker. And continuously verifying, and iterating to obtain the optimum smoothness value x, so as to finish a more accurate smoothness-controllable coffee roasting state identification method and assist a coffee baker to roast better.

The above embodiments are merely illustrative of the technical ideas and features of the present invention, and the purpose thereof is to enable those skilled in the art to understand the contents of the present invention and implement the present invention, and not to limit the protection scope of the present invention. All equivalent changes and modifications made according to the spirit of the present invention should be covered within the protection scope of the present invention.

Claims

1. A coffee bean heating rate smooth and controllable and roasting state automatic identification method is characterized in that:

the coffee machine completes Socket communication with the mobile phone through the mobile phone and the distribution network thereof, and communicates with a back-end server through a router;

the temperature module in the coffee machine detects the temperature, sends the bean body temperature to the back-end server and the mobile phone end, records the bean body temperature in the temperature linked list, calculates the temperature rise rate according to the bean body temperature in the temperature linked list, establishes the temperature rise rate linked list, and judges the roasting state of the coffee beans according to the temperature rise rate linked list;

selecting bean species to obtain smoothness, baking state temperature information and bean species temperature rising characteristics before baking starts, then selecting whether to start an automatic identification mode, if so, obtaining the optimal smoothness of the current temperature rising rate by using a greedy algorithm, and obtaining the global optimal smoothness through the local optimal smoothness each time;

the length of the temperature linked list is adjustable, and when the number of the received temperature values reaches the length of the linked list, the temperature value of the queue head is popped up every time a temperature value is transmitted;

the temperature rise rate R is calculated by adopting the tail end temperature L, the third tail end temperature L3 and the head temperature F in a temperature linked list, and the formula is as follows: r =60(2L-F-L3)/(x +3), where x is the smoothness of the setting.